Sign up to save your podcasts
Or
Share Why Agentforce Is a Game Changer for Small Businesses
Share to email
Share to Facebook
Share to X
Share to email
Share to Facebook
Share to X
Or
Why Agentforce Is a Game Changer for Small Businesses
Today on the Salesforce Admins Podcast, we talk to Daniel Peter, Chief Technology Officer at Petaluma Creamery. Join us as we chat about how he manages cheese wheels with custom objects and how Salesforce and AI can level the playing field for SMBs.
You should subscribe for the full episode, but here are a few takeaways from our conversation with Daniel Peter.
Modernizing business processes at a historic creamery
Daniel gave out some of the best swag at Dreamforce—free cheese samples. As a Salesforce MVP Hall of Fame member, he’s held a wide range of roles on the platform, but none have been quite as delicious as his current gig as the Chief Technology Officer at Petaluma Creamery. You could say he’s the big cheese for digital transformation.
The creamery is a 115-year-old business capable of producing over 140,000 pounds of cheese per day. It’s an old-school business, and that means he inherited several old-school business processes. With so many manual processes, Daniel had to move fast and focus on the biggest wins first.
Digital transformation priorities for SMBs
So how did Daniel take his business processes from aged Gouda to fresh mozzarella? He started with the basics: getting the cheddar through the door. In other words, simplifying the ordering process.
Like a lot of SMBs, the creamery’s system dated back to a time when you could just throw more people at a logistics problem. A sales or delivery person would talk through the order with the customer, fill out a paper form, and then do some unit conversions before they could enter the data into a database. It was time-intensive, labor-intensive, and introduced all kinds of opportunities for mistakes.
Daniel quickly built an order system in Salesforce that saves time, does all of the conversions on the backend, and makes it easier for his users to find the product they’re looking for. The creamery is also able to track all sorts of data about the cheese-making process, like where ingredients come from and how they were stored, which is crucial for getting a certified-organic label.
Why Agentforce levels the playing field for SMBs
A common misconception is that AI tools are reserved for huge corporations with the technical resources to implement them. However, as Daniel explains, affordable tools like Agentforce are leveling the playing field for SMBs.
Looking forward, he’s aiming to implement several agents that will streamline the creamery’s business processes:
If all of this sounds exciting to you, be sure to check in with Daniel at TDX to see what he’s built. And sample some delicious dairy.
Make sure to listen to the full episode for more from Daniel, and don’t forget to subscribe to the Salesforce Admins Podcast to catch us every Thursday.
Podcast swag
Learn more
Admin Trailblazers Group
Social
Full show transcript
Mike:
Hold onto your curds folks. Today’s episode is un-brie-lievable. We’re chatting with Daniel Peter, CTO at Petaluma Creamery, where they’re proving that even in a world full of mozzarella sticks and spreadsheets, Salesforce can be the big cheese. Now today, Daniel’s going to dish on how he’s gone from QuickBooks and paper trails to a cheddar-rific automation and including some AI agents. So if you’ve ever wondered what it’s like to manage wheels of cheese with custom objects, or if curd record type is a thing, then this is your jam. I mean, cheese, let’s melt into it. Let’s get Daniel on the podcast.
So Daniel, welcome to the podcast.
Daniel Peter:
Hey, thanks for having me.
Mike:
I know it’s been a while since we’ve had you on, it was Talking Community and Twitter back when it was called Twitter.
Daniel Peter:
Yeah, I did look up that podcast, it was 2017.
Mike:
Oh man, that was a thousand years ago.
Daniel Peter:
Yup.
Mike:
For those people that aren’t familiar with who Daniel Peters is, can you give us a quick update and kind of let everyone know what you do and how you got started in the ecosystem?
Daniel Peter:
Yeah. I got started way back in 2009 as a developer, and I’ve done a lot of things. I’ve worked for Salesforce customers, ISV Partners, did a stint at Robots and Pencils where I ran a SI Salesforce practice, done a lot of consulting work on the development side for ISV, and actually even for Salesforce. But my most recent venture here, I’m back in the customer seat. So I’m the chief technology officer here at the Petaluma Creamery, and we’re having a lot of fun relaunching this place on Salesforce. So that’s what I’m up to now. And I’m also a Salesforce MVP Hall of Famer.
Mike:
Aha. There, snuck it in just by the way, buried the lead.
Daniel Peter:
Yeah, I got that in.
Mike:
I think we met when you were at the nonprofit.
Daniel Peter:
I don’t remember which nonprofit that was.
Mike:
I thought it was the Pencils one anyway.
Daniel Peter:
Oh, Robots and Pencils, yeah.
Mike:
Yeah, Robots and Pencils. Sorry. I knew there was pencils and paper. Pens and papers and [inaudible 00:02:42].
Daniel Peter:
Yeah. So Robots and Pencils is a Salesforce nonprofit partner, but only on the higher ed side, not on the NGO side. And I know that whole organization has changed with Education Cloud and all of that. So yeah, that threw me off for a second. But yeah, technically we were, yeah, we were a nonprofit partner on the education side.
Mike:
Okay. So I know my friends are always like, every time I see customers, I’m like, oh yeah, I think that’s a Salesforce customer. So now you work at a creamery, and I ran into you at Dreamforce. You were handing out cheese, which as a Midwesterner, that’s exactly how you get straight to my heart is like you either hand somebody from the Midwest, a piece of meat or a piece of cheese.
Daniel Peter:
Yeah, I think I’m the only one with cheese swag [inaudible 00:03:35].
Mike:
Yeah, I think you are. Tell me a little bit about what being the CTO at a creamery is like, and what are some of the challenges? I mean, what do you manage the curds in Salesforce? What do you do? Is there a curd record type?
Daniel Peter:
You know what there basically is, I mean, it sounds funny, but all that stuff is really important to track for all these different compliance reasons. If you want to be organic certified, you have to show the milk coming in, that batch that you made out of it, butter, cheese, yogurt. You have to show the ingredients that went in. And all of that really, I was at Canandy and we built a manufacturing ERP for Del Monte, and so I knew Salesforce was up to the task of tracing all the ingredients and manufacturing of food. So yeah, we do store all of that in Salesforce.
Mike:
So how were they doing it before you started?
Daniel Peter:
It was pretty fragmented. The main system we had for our kind of the closest thing to ERP was QuickBooks Desktop. And they used QuickBooks Desktop for 20 years, and it was the big manufacturing edition. He actually was doing a lot of business here. There was years when he did close to 50 million a year and all that ran through QuickBooks, and it did have some inventory capabilities, but it’s pretty terrible to use. And then there was just a lot of paper-based, like the order forms were on paper. They would take inventory on paper and then put it into an Excel spreadsheet.
Mike:
Did you have to do your compliance stuff on paper too?
Daniel Peter:
Yeah, there’s rooms here full of paper files. I mean, so it’s kind of an old school industry, and I think a lot of the people in it, they know cows, they know how to make cheese, but technology isn’t something that comes naturally to them. They’re much more comfortable writing things on paper.
Mike:
Well, even the technology of cheese has changed too. I mean, I know where I live in Iowa, we have a creamery that’s about 20 miles south of us, and it’s run by Amish, and that’s always a big trip because you got to head down there to get some Amish cheese and bread. Oh, the bread is amazing. Let me know when you work for a bakery, by the way. But the technology of cheese, because they have the open windows, that’s changed. But ironically, I think sometimes the technology that manages the processes, or in your case maybe just the data, just gets left to pencil and paper because it’s what we know.
Daniel Peter:
Yeah. And an interesting story here is that the pencil and paper at one point was an economical way to do things. I mean, that actually created a lot of jobs for people here. This creamery has been here for over 115 years, and back then you would just throw more people at the problem. But that’s kind of the changing paradigm that we’re in is the city of Petaluma has grown up around this creamery. This is a Bay Area kind of city, and property prices are through the roof, labor costs are through the roof, utilities are through the roof. You can’t really exist anymore in a world where you’re just throwing people at the problem. It’s way too expensive and your margins will disappear quickly.
Mike:
Well, I mean there’s really, to be realistic, there’s only so much you could sell that block of cheese for. And you can’t pay 15 people to live in a high dense area with all of that cost just to move paper around. If a one pound block of cheese is going to end up costing $400 just to make everybody’s living wage, right?
Daniel Peter:
Exactly. And that’s why most of the creameries have moved out of California, out of state into the lower cost of living. Yeah, that block of cheese is for commodity cheese it’s kind of a race to the bottom who can make it for the lowest cost. But we’re actually a little bit more of a differentiated product. So our cheese isn’t competing for lowest price. We’re actually competing more on quality and kind of a best cost. It’s a pretty good price, but also really good quality.
Mike:
Yeah. So what was some of the first things you tackled to move off of pen and paper and give a little more structured process around what you’re doing and integrate Salesforce?
Daniel Peter:
Yeah. Well, the very first thing was to stabilize just the IT here, which was get off of T-One lines if anyone knows what those are and get on fiber optic, get rid of on-premise servers. There was an exchange server, an Outlook server, a QuickBooks server. Somebody could walk by and accidentally turn it off or it might just die, and then all your systems are down. So getting good internet and getting on the cloud, that was the very first thing. But the first interesting thing we did in Salesforce was actually I wanted to make it, we need to stay alive here. We need to make money. So I wanted to make it super easy to take orders, get the cash in the door, the order to cash process.
Mike:
Sure.
Daniel Peter:
So QuickBooks was so painful. To put an order in QuickBooks, you had to actually do math either in your head or on a calculator. You had to convert, like if somebody ordered a case of cheese, you had to know, oh, those are eight ounce pieces and there’s 14 in a case, so that’s seven pounds. Okay, so they say one case. So we have to convert that to seven pounds in your head to put it into QuickBooks. And then you also had to do things like memorize the whole cheese hierarchy. So like C:J for Jack, M for Monterey, 8 for eight ounce. That’s how you punch it into QuickBooks. And yeah, you basically had to have a computer in your head just to enter an order into the computer, and it was super painful.
So we couldn’t get orders into the system quickly, and we can’t just… We’d have to highly train people and it’s highly error-prone to put orders in and people would call on the phone or you’d get a written order, and then you’d have to convert that in QuickBooks. So I said, what we need to do is just bury QuickBooks on the backend. Let’s build a really nice, basically an order entry screen in Salesforce. So we built it exactly like we want it for our business in Salesforce. So if customers want to order by the pound, by the piece, by the case, Salesforce does all the conversions, ultimately converts it back into pounds in QuickBooks, what’s integrated with QuickBooks.
But it lets you enter it in a very human friendly way, very nice auto complete. I mean, you can basically just start to type one letter of Y for yellow cheddar, and yellow cheddar just pops up. The orders almost right themselves now, and it’s very easy to train new people how to put orders in the system. So that was a huge-
Mike:
Oh, that’s good.
Daniel Peter:
Yeah. That was a huge win.
Mike:
Did you also stand up an external facing way to take orders too, to speed that up, or is that on the roadmap?
Daniel Peter:
That’s still on the roadmap. We do have, so 99% of our business comes from wholesale to stores.
Mike:
Oh, okay.
Daniel Peter:
We do have a small segment of direct-to-consumer, so you can go to springhillcheese.com and order online. Right now that’s just a Squarespace e-commerce website, and that works well, but we want to do some fancier things. So we might look into some Salesforce offerings. There’s obviously things like Commerce Cloud, but there’s also some ISVs have built some apps on Salesforce to sell. So we’ll look into some of those. Right now during this relaunch and focusing on staying alive here in the short term, really wholesale is what we’re focusing on. But there’s a big opportunity in direct-to-consumer. We ship cheese in a kind of an insulated cooler. Believe it or not, it actually ships really well throughout the US, and a lot of people do order. That’s on the roadmap, actually. But the tricky thing is the stores might not necessarily… The stores, a lot of them are still old school and the buyers still kind of want to call you. So even if we did give them a way to enter their own orders, not everybody would adopt it.
Mike:
Yeah, they’d still call you.
Daniel Peter:
Yeah, there’s still a human element there that we need to preserve. But obviously you can start to dream about how AI can listen in on your conversation and create the order for you as you’re talking, things like that.
Mike:
Quiet, AI is listening right now, Daniel, it’s taking my cheese order. And I mean, that completely makes sense. You want, for a business, you want to send truckloads of cheese, not shoe boxes of cheese. Because even a consumer, somebody like me thinks they’re ordering a lot of cheese. Well, compared to a grocery store or a supermarket chain that’s ordering for five stores to stock 10 coolers, focusing on wholesale totally makes sense. So keeping that human element there, you can always go to the consumer, but then you’re shipping probably smaller chunks of items. So you really need the volume first.
Daniel Peter:
Yeah. If people want to order just like a piece of cheese, they end up paying more for the shipping than the cheese costs. But we do free shipping over $100. That seems to be a good sweet spot. If people want to order, then it’s like say it costs $100 worth of cheese and we pay the $30 shipping, that kind of thing. So there is a market for the direct-to-consumer when you want to purchase a little bit bigger order, it kind of makes more sense. But yeah, this place can make 140,000 pounds of cheese a day, and that’s one of its-
Mike:
Wow.
Daniel Peter:
… images we need to leverage. So yeah, definitely moving by the palette or by the semi-truck load is how we can make money here.
Mike:
Okay. So you said something, and we talked at Dreamforce that I think a lot of small or medium-sized businesses are faced with, which is they all, “I’m too small for AI.” Like AI is really for the big players like the Fortune 500 or the Fortune 100 companies need AI. But you’re just a creamery, not just, I mean, you’re a small creamery out in California, and before I pressed record, I mean you were at Dreamforce, you saw all the Agentforce stuff, you saw where AI is going. What are your plans for AI?
Daniel Peter:
Yeah, I think the story we hear a lot is about AI replacing jobs. There’s a lot of fear I guess, in the job market, both on the developer side, admin, but also basically any kind of a white collar worker that might be doing any type of work is my job going to be replaced by AI? But I think there’s all these small businesses that really have a chance to have a renaissance because of AI. So instead of losing jobs, it’s more like AI can enable them to save themselves, to save an industry or to save a small-medium business. We’ve seen, unfortunately, a lot of… Fortunately, or unfortunately, I guess there’s been a lot of disruption over decades. A lot of mom and pop stores can’t afford to be around because of places like Amazon. It’s kind of changed the world, but Amazon’s a great service too.
I mean, you can get anything you want in a couple of days, so you could say it’s fortunate or unfortunate. But I think people like to have some of these old school industries around. It kind of makes life more interesting to be able to go visit a historic place instead of everything just coming from a centralized factory. Yeah, I think the story for these smaller industries is, hey, you can actually compete. AI levels the playing field a bit now. You can actually afford AI, and it’s pretty easy to use with tools like Salesforce. So you can actually make your business reinvent your business model, to be efficient enough to be profitable, this new paradigm. So that’s the story we have here, and that’s why it’s exciting to me here.
And so I think, yeah, small business, you need a little bit of tech talent. You might have to find, like my cousin that owns this creamery came and found me to help them out with that. You might have to find somebody to help you, but it’s really easier now than it’s ever been. And I think people are pretty comfortable now doing things like writing customer service emails with ChatGPT that we’ve achieved a certain threshold where the fear is kind of going away and the trust is starting to come into play. And so it’s just a matter of, okay, well, how can I implement this now at scale for my business, to where I’m focusing on the things I’m really good at? Like my vision for the business and interacting with the customers that come in the door while AI is doing all the grunt work in the background for you, and at a pretty low cost.
Mike:
So you envision at your creamery using AI as kind of a scalable factor? Like you can scale one person, as opposed to how our conversation began, which is you would just throw people at it. AI kind of gives you that ability to throw people at it, except you’re just throwing an agent at it, right?
Daniel Peter:
Yeah. It’s sort of what the cloud did for computing when we needed elasticity. When you got a Black Friday and you got a million requests all at once and your server went down, and the cloud, kind of solve that. I think AI solves the people elasticity problem. Certain times you have a lot of orders coming in, and do you really want to have 10 people on staff all the time when they’re mostly sitting around? But then once in a while, all 10 phones are ringing. AI can solve that pretty easily.
Mike:
So you saw a lot at Dreamforce. I mean, we’ve rolled out so many things, it’s hard to remember everything, but I mean, even my team is sitting back trying to build examples for agents and prompting. As a small business. And I think it’s fair to say that where are you looking at deploying, or hey, this could be a good use case for Agentforce, or really thinking about how we can scale there.
Daniel Peter:
Yeah, we’ve got a few great ones. I think they’re all similar level of importance. So this will be in no particular order, but one of them is an agent to take orders. It’s very smart. So it would be something we’d use internally. Maybe even a delivery driver in the field might dictate it from their phone, or we could use it right in the office, but instead of having to type things into the order entry screen, we can just talk to it, talk to the agent. Ideally even with our voice and just say, “Hey, for Safeway, number 572 in San Francisco, they called in, they want to duplicate their order, but they want to add in an extra case of this.”
And it has all that context already of what they ordered before. So you don’t have to go look it up and try to clone it and add to it, and it can just crank out that order. And it can say, is this what you meant? And you can say, yeah, so there’s that one we want to build. There’s also one that’s more customer focused, and that would be, we get a lot of calls, emails for random customer service requests that are… They’re along the lines of refunds, credits, changes, questions, things like this. And again, we have all that context. We had QuickBooks for 20 years, so we have 20 years of their order history imported into Salesforce.
Mike:
Wow. Wow. That’s huge.
Daniel Peter:
Yeah. So we don’t have a data problem. Actually we have plenty of good data here, but these things eat up a lot of time. So people call in and they’ll send a very vague request. They’ll say something like, we need three replacement pieces of yellow cheddar. And so we make so many different types of yellow cheddar. And then once you figure out what they need, how do you handle that? Do you add them to their next order? Do you create a credit memo? These kinds of things. So you can imagine an agent knows everything about the customer’s history. It fills in the blanks on these vague requests, and it does the action. Whether it’s adding something onto the order, issuing a credit, that kind of thing.
And if it needs to, let’s say emails the channel that could clarify through an email. Maybe if it’s a vague request and we have context, but the responses are still ambiguous, there may be one thing we need to clarify. It can do that agentic loop and iterate until the customer’s happy with the outcome. So that’s another one. And then the final one that I was thinking would be really fun to deploy, it’s kind of a… I haven’t thought of a name for it yet, but basically be kind of a cheese agent that really knows everything about all of our products. So we’ve got all the order history, but what we also have is a super rich library of unstructured data here.
So there was a world-class lab here on site for 20 years. It’s one of the three blocks that this place is on, and it had all kinds of testing equipment. And they cranked out so many analysis, studies of all the products. So anything, any random long tail questions somebody might want to ask about dairy products, cheese, butter, powdered products, any of that, we have the answer to sitting in these unstructured documents. Anything from PowerPoint presentations to Excel to PDFs, and I can just imagine putting that into data cloud and being able to produce a really good answer.
So I think probably a way better answer than a customer service rep is ever going to be able to come up with, because how are they going to be able to do all that research?
Mike:
Yeah. It’d be just based on their knowledge. And then it’s a lottery based on whoever’s working that day.
Daniel Peter:
Yeah. So if people want to know salt content and this or that, or do you use this type of rennet that had this particular genetically modified organism or not? Or how many days out of the year are your cows eating grass? Or how long is this product good in our refrigerator at 38 degrees? We can answer just infinite number of these long tail questions. So I think that one would be not just an efficiency gainer, but actually would increase our customer satisfaction by giving them not only the information they want, but these generative AI responses have a very nice friendly tone to them also. So you’d get a great answer and it would also make you feel good, because it’s so nice to you. And so I think that would be huge for customer satisfaction.
Mike:
Yeah, I mean, I was even thinking the consumer route, once you go that way is a lot of people, when you go to the butcher, it’s the same way too. You have a general idea of what you could do with this, but especially with cheeses, the right or wrong cheese can make or break a dish. And even combinations of cheese. I like watching the Food Network, and they’ll make a macaroni and cheese, and it won’t just be one cheese, it’ll be five. And it’s funny because my mom will be like, “How do they know those five tastes good together?” I’m like, “Well, mom, they’re trained chefs.” They’re not like you and me where it’s trial and error. Let’s add this in, and it totally makes it a mess. But you could even be that same way too. Oh, if you’re ordering these three cheeses, you could order these other two. And they pair really well, all five of them together. And then they’re ordering more cheese.
Daniel Peter:
Yeah, totally. And yeah, I’ve experienced that. So we have this four-way shred blend we do for the Kings, Sacramento Kings Stadium.
Mike:
Oh, wow.
Daniel Peter:
And I don’t know what it is about that mix, but those four cheeses together, they synergize. And I go out and sample it when they’re making it, and it’s way better than… You could have one, two, or three of the cheeses. But when you put all four together, it’s like something magic happens with the combination-
Mike:
[inaudible 00:26:26], yeah.
Daniel Peter:
… cheese and some spicy cheese, and maybe there’s some mozzarella in there or something. But the way they… Yeah. So there’s definitely something to getting that magic mix right. And I think that, yeah, that’s something really fun, that’d be a fun AI problem to solve.
Mike:
That would also, you think about that. It would help educate your consumer, but it also helps sell more cheese. The good salesperson doesn’t just sell what the customer wants. It also sells what the customer doesn’t know they want.
Daniel Peter:
Yeah.
Mike:
Right? Like I know I’m going to come to you and just buy some cheddar or some Parmesan, but I’m buying these two. Did you know if you mix this with it, or here’s five other recipes you could use it for. So it’s always kind of… And again, these are all things a small little creamery is here’s how we can add scalable people on with agents that more than offset the cost of it. And you’re not a thousand person company or a 10,000 person company that’s doing it, right? But you can act like you have a thousand people.
Daniel Peter:
Yeah. So a lot of the stuff you’re mentioning is things that really good salespeople are good at, and really good salespeople are hard to find. And we have some here, but yeah, they don’t scale. So yeah, that upsell opportunity, I mean people… That actually it’s a win-win. We make more money and they actually walk away feeling like they had a way better experience. They got to try this new thing they wouldn’t have tried otherwise, something like that. So yeah, that’s pretty exciting when you can have AI fill in for that lack of top sales talent like that.
Mike:
So I guess we’ll end on a fun note. What is something that you saw at Dreamforce that you can’t wait to get your hands on?
Daniel Peter:
Oh boy. Well, I know this is the admin podcast, but-
Mike:
That’s okay. You can talk about any of our products.
Daniel Peter:
I’m pretty excited about the vibe coding. Basically, I’ve been a big fan of generative AI just as an easy button for everything and seeing how good it can write your code for you now. I was a little skeptical at first just because I’ve been a developer for so long, but when I saw it in action, I realized that’s going to be the next game changer for developers. So I actually, I haven’t been using it much, and I want to start using it now, and it’ll actually help to build some of these agents that I’m talking about more quickly. So that’s huge. I mean, I’m wondering, we’ve seen generative AI disrupt so many different tools that we use.
It’s what people use now for, we saw the students using it to write their term papers for them, and it’s just gone the gamut of things. But now actually it’s smart enough to be able to write code in quality, and it’s a really high-level skill that’s going to be a huge time saver for developers.
Mike:
Yeah, I think, not to belabor the point, and I certainly have never written code, there’s always, oh, it’s going to take our jobs. Well, technology’s evolved since forever, but it’s always taken the mundane parts of our job. And that’s how I see a lot of… Like I sat in the developer keynote and watched it. To me coding, I think they demoed building a menu, I don’t know, Daniel, you tell me. If you had to sit down and code something, I have to believe that coding the menu, the mundane part wasn’t the part you were most excited to work on.
Daniel Peter:
No.
Mike:
But if AI can do it and then it can be like, okay, so here’s the 10% I can’t do, that’s probably the 10% you really wanted to work on anyway.
Daniel Peter:
It’s so true. And that’s actually been what Salesforce has been good at since the very beginning. You go back to when you would create a custom field, you wouldn’t have to go put it in the database and put it in each TML page. It just showed up everywhere automatically. And that’s all you wanted was a custom field. You didn’t want to have to write all the tedious code to wire it from the bottom to the top of the stack. So yeah, it’s just the next logical evolution of Salesforce making it to where you just get to do what you want to do and it handles the mundane things.
Mike:
Yeah.
Daniel Peter:
Yeah. Salesforce was an early adopter to that lower code way of doing things. And it’s exciting that now it’s come all the way to actually pro code can be written with your voice.
Mike:
I mean, there’s always a need for both. I think you’re kind of a car guy. I saw a stat that like 94% of the automobiles produced are automatic transmissions now. Well, that’s still 6% that aren’t, and is that good or bad? Well, I mean, some people that really like three pedals, oh, that’s horrible, why are they making all these automatic cars? They’re taking the fun away and this. Yeah, I mean, to be fair, you drive a 50-year-old manual clutch car, it’s a lot different than when you get in your car. And I mean, some of them don’t even have a gear shift now, some of them, you just press a button to start and press a button to put it in gear.
But then also the experience of driving. You can focus on the things that really help you prevent getting in a car crash, as opposed to just making the car go the road. And as a car guy, it’s two very different things. One’s not better than the other, it’s just less weight on your mind.
Daniel Peter:
Yeah. Yeah, that’s a cool analogy. I like that.
Mike:
So I mean, they’ll probably always still make manual transmission cars. If not, I’ll just keep mine going. And that doesn’t make them any less, it just means that there’s less tediousness for you to do to make the car go down the road so you can focus on some of the other things. Perspective, right?
Daniel Peter:
Yeah, no, I like that perspective on it. It makes a lot of sense. So yeah, don’t fear the loss of having to figure out how to use your clutch and your gear shift, just embrace it.
Mike:
I mean, I still think everybody should learn how to drive a manual transmission car. It does make you a better driver. I guess the opposite of that is I imagine learning how to code would make you a better admin too. So the reverse of everything has to hold true.
Daniel Peter:
Yeah. I know it gets into a pretty deep discussion about maybe how long until code totally goes away, at least from our perspectives, creating business software on Salesforce. I think I could see that even happening. And then it’s like, well, you don’t even have to drive the automatic, it’s more like the Waymo or something.
Mike:
Even then somebody has to write the software that makes it so that you don’t have to write code.
Daniel Peter:
Yeah, there’ll always be at some level. Yeah, I think there’ll always have to be. It just keeps moving further away from the people that are familiar with the business and more down to that platform level.
Mike:
Yeah. I guess I look at it much in the same of an engineering. The people that engineer and make automatic transmissions, I mean, they make automatic transmissions that can hold thousands of horsepower. And these engineers are incredibly intelligent. That doesn’t mean that the mechanic is any less of an engineer. It just means that the engineering skill moved in a different part of the business. Much like the coding skill or the non-coding skill.
Daniel Peter:
Yeah. And I think that’s good. I mean, I think people are scared of it, but I think businesses should be in business to focus more on how they can basically make their customers happier. Not so much in having to do all the tedious behind the scenes things to make it happen.
Mike:
Right. Well, Daniel, we’ve covered a lot today. Cheese, manual transmissions, business process, AI. I think if I had a wheel, we would’ve spun it and hit every corner of the wheel.
Daniel Peter:
Yeah. Well, I’m excited to keep building out these agents we’re talking about. And I’ll keep everybody in the ecosystem updated through different means. I’ve been doing some different articles and things like that, posting on LinkedIn, and so excited to get more and more of these cheese agents rolled out, and get us scaled up without having to eat up all of our margins and then some.
Mike:
Oh, look at you. Eat up all of your margins. I see where you’re going with that. Good job. Well, Daniel, I look forward to, we got TDX coming up in April of next year. Hopefully we can see one of your cheese agents at TDX.
Daniel Peter:
Yeah, definitely. Yeah, I kind of wish that I had more to show off a Dreamforce this year, but I’ve been wearing a lot of hats here as CTO of a creamery, so we need to get this stuff rolled out like yesterday.
Mike:
Yeah. Well, I mean, the cheese has got to make the money before you can buy more software. So I think you got your priorities straight.
Daniel Peter:
Yeah.
Mike:
Yeah. And you make good cheese too.
Daniel Peter:
Oh, thank you. Yeah, I love it. It’s a little dangerous working here.
Mike:
I could only imagine.
Daniel Peter:
All the cheese you can eat.
Mike:
Can only imagine. Well, Daniel, thanks so much for coming on the podcast, and I look forward to seeing your cheese agents in the future.
Daniel Peter:
Thanks for having me, Mike. This was a lot of fun.
Mike:
Well, that was cheddar than expected. So big thanks to Daniel Peter for cracking open the cheese vault and sharing how Salesforce is helping Petaluma Creamery brie-come more efficient, more scalable, and still delicious. Yes.
Remember, digital transformation, including Agentforce, isn’t just for the big cheeses, even small creameries can scale with AI and Salesforce. No provolone required. So spread the word, and until next time, we’ll see you in the cloud.
The post Why Agentforce Is a Game Changer for Small Businesses appeared first on Salesforce Admins.
View all episodes
By Mike Gerholdt
4.7
201201 ratings
Today on the Salesforce Admins Podcast, we talk to Daniel Peter, Chief Technology Officer at Petaluma Creamery. Join us as we chat about how he manages cheese wheels with custom objects and how Salesforce and AI can level the playing field for SMBs.
You should subscribe for the full episode, but here are a few takeaways from our conversation with Daniel Peter.
Modernizing business processes at a historic creamery
Daniel gave out some of the best swag at Dreamforce—free cheese samples. As a Salesforce MVP Hall of Fame member, he’s held a wide range of roles on the platform, but none have been quite as delicious as his current gig as the Chief Technology Officer at Petaluma Creamery. You could say he’s the big cheese for digital transformation.
The creamery is a 115-year-old business capable of producing over 140,000 pounds of cheese per day. It’s an old-school business, and that means he inherited several old-school business processes. With so many manual processes, Daniel had to move fast and focus on the biggest wins first.
Digital transformation priorities for SMBs
So how did Daniel take his business processes from aged Gouda to fresh mozzarella? He started with the basics: getting the cheddar through the door. In other words, simplifying the ordering process.
Like a lot of SMBs, the creamery’s system dated back to a time when you could just throw more people at a logistics problem. A sales or delivery person would talk through the order with the customer, fill out a paper form, and then do some unit conversions before they could enter the data into a database. It was time-intensive, labor-intensive, and introduced all kinds of opportunities for mistakes.
Daniel quickly built an order system in Salesforce that saves time, does all of the conversions on the backend, and makes it easier for his users to find the product they’re looking for. The creamery is also able to track all sorts of data about the cheese-making process, like where ingredients come from and how they were stored, which is crucial for getting a certified-organic label.
Why Agentforce levels the playing field for SMBs
A common misconception is that AI tools are reserved for huge corporations with the technical resources to implement them. However, as Daniel explains, affordable tools like Agentforce are leveling the playing field for SMBs.
Looking forward, he’s aiming to implement several agents that will streamline the creamery’s business processes:
If all of this sounds exciting to you, be sure to check in with Daniel at TDX to see what he’s built. And sample some delicious dairy.
Make sure to listen to the full episode for more from Daniel, and don’t forget to subscribe to the Salesforce Admins Podcast to catch us every Thursday.
Podcast swag
Learn more
Admin Trailblazers Group
Social
Full show transcript
Mike:
Hold onto your curds folks. Today’s episode is un-brie-lievable. We’re chatting with Daniel Peter, CTO at Petaluma Creamery, where they’re proving that even in a world full of mozzarella sticks and spreadsheets, Salesforce can be the big cheese. Now today, Daniel’s going to dish on how he’s gone from QuickBooks and paper trails to a cheddar-rific automation and including some AI agents. So if you’ve ever wondered what it’s like to manage wheels of cheese with custom objects, or if curd record type is a thing, then this is your jam. I mean, cheese, let’s melt into it. Let’s get Daniel on the podcast.
So Daniel, welcome to the podcast.
Daniel Peter:
Hey, thanks for having me.
Mike:
I know it’s been a while since we’ve had you on, it was Talking Community and Twitter back when it was called Twitter.
Daniel Peter:
Yeah, I did look up that podcast, it was 2017.
Mike:
Oh man, that was a thousand years ago.
Daniel Peter:
Yup.
Mike:
For those people that aren’t familiar with who Daniel Peters is, can you give us a quick update and kind of let everyone know what you do and how you got started in the ecosystem?
Daniel Peter:
Yeah. I got started way back in 2009 as a developer, and I’ve done a lot of things. I’ve worked for Salesforce customers, ISV Partners, did a stint at Robots and Pencils where I ran a SI Salesforce practice, done a lot of consulting work on the development side for ISV, and actually even for Salesforce. But my most recent venture here, I’m back in the customer seat. So I’m the chief technology officer here at the Petaluma Creamery, and we’re having a lot of fun relaunching this place on Salesforce. So that’s what I’m up to now. And I’m also a Salesforce MVP Hall of Famer.
Mike:
Aha. There, snuck it in just by the way, buried the lead.
Daniel Peter:
Yeah, I got that in.
Mike:
I think we met when you were at the nonprofit.
Daniel Peter:
I don’t remember which nonprofit that was.
Mike:
I thought it was the Pencils one anyway.
Daniel Peter:
Oh, Robots and Pencils, yeah.
Mike:
Yeah, Robots and Pencils. Sorry. I knew there was pencils and paper. Pens and papers and [inaudible 00:02:42].
Daniel Peter:
Yeah. So Robots and Pencils is a Salesforce nonprofit partner, but only on the higher ed side, not on the NGO side. And I know that whole organization has changed with Education Cloud and all of that. So yeah, that threw me off for a second. But yeah, technically we were, yeah, we were a nonprofit partner on the education side.
Mike:
Okay. So I know my friends are always like, every time I see customers, I’m like, oh yeah, I think that’s a Salesforce customer. So now you work at a creamery, and I ran into you at Dreamforce. You were handing out cheese, which as a Midwesterner, that’s exactly how you get straight to my heart is like you either hand somebody from the Midwest, a piece of meat or a piece of cheese.
Daniel Peter:
Yeah, I think I’m the only one with cheese swag [inaudible 00:03:35].
Mike:
Yeah, I think you are. Tell me a little bit about what being the CTO at a creamery is like, and what are some of the challenges? I mean, what do you manage the curds in Salesforce? What do you do? Is there a curd record type?
Daniel Peter:
You know what there basically is, I mean, it sounds funny, but all that stuff is really important to track for all these different compliance reasons. If you want to be organic certified, you have to show the milk coming in, that batch that you made out of it, butter, cheese, yogurt. You have to show the ingredients that went in. And all of that really, I was at Canandy and we built a manufacturing ERP for Del Monte, and so I knew Salesforce was up to the task of tracing all the ingredients and manufacturing of food. So yeah, we do store all of that in Salesforce.
Mike:
So how were they doing it before you started?
Daniel Peter:
It was pretty fragmented. The main system we had for our kind of the closest thing to ERP was QuickBooks Desktop. And they used QuickBooks Desktop for 20 years, and it was the big manufacturing edition. He actually was doing a lot of business here. There was years when he did close to 50 million a year and all that ran through QuickBooks, and it did have some inventory capabilities, but it’s pretty terrible to use. And then there was just a lot of paper-based, like the order forms were on paper. They would take inventory on paper and then put it into an Excel spreadsheet.
Mike:
Did you have to do your compliance stuff on paper too?
Daniel Peter:
Yeah, there’s rooms here full of paper files. I mean, so it’s kind of an old school industry, and I think a lot of the people in it, they know cows, they know how to make cheese, but technology isn’t something that comes naturally to them. They’re much more comfortable writing things on paper.
Mike:
Well, even the technology of cheese has changed too. I mean, I know where I live in Iowa, we have a creamery that’s about 20 miles south of us, and it’s run by Amish, and that’s always a big trip because you got to head down there to get some Amish cheese and bread. Oh, the bread is amazing. Let me know when you work for a bakery, by the way. But the technology of cheese, because they have the open windows, that’s changed. But ironically, I think sometimes the technology that manages the processes, or in your case maybe just the data, just gets left to pencil and paper because it’s what we know.
Daniel Peter:
Yeah. And an interesting story here is that the pencil and paper at one point was an economical way to do things. I mean, that actually created a lot of jobs for people here. This creamery has been here for over 115 years, and back then you would just throw more people at the problem. But that’s kind of the changing paradigm that we’re in is the city of Petaluma has grown up around this creamery. This is a Bay Area kind of city, and property prices are through the roof, labor costs are through the roof, utilities are through the roof. You can’t really exist anymore in a world where you’re just throwing people at the problem. It’s way too expensive and your margins will disappear quickly.
Mike:
Well, I mean there’s really, to be realistic, there’s only so much you could sell that block of cheese for. And you can’t pay 15 people to live in a high dense area with all of that cost just to move paper around. If a one pound block of cheese is going to end up costing $400 just to make everybody’s living wage, right?
Daniel Peter:
Exactly. And that’s why most of the creameries have moved out of California, out of state into the lower cost of living. Yeah, that block of cheese is for commodity cheese it’s kind of a race to the bottom who can make it for the lowest cost. But we’re actually a little bit more of a differentiated product. So our cheese isn’t competing for lowest price. We’re actually competing more on quality and kind of a best cost. It’s a pretty good price, but also really good quality.
Mike:
Yeah. So what was some of the first things you tackled to move off of pen and paper and give a little more structured process around what you’re doing and integrate Salesforce?
Daniel Peter:
Yeah. Well, the very first thing was to stabilize just the IT here, which was get off of T-One lines if anyone knows what those are and get on fiber optic, get rid of on-premise servers. There was an exchange server, an Outlook server, a QuickBooks server. Somebody could walk by and accidentally turn it off or it might just die, and then all your systems are down. So getting good internet and getting on the cloud, that was the very first thing. But the first interesting thing we did in Salesforce was actually I wanted to make it, we need to stay alive here. We need to make money. So I wanted to make it super easy to take orders, get the cash in the door, the order to cash process.
Mike:
Sure.
Daniel Peter:
So QuickBooks was so painful. To put an order in QuickBooks, you had to actually do math either in your head or on a calculator. You had to convert, like if somebody ordered a case of cheese, you had to know, oh, those are eight ounce pieces and there’s 14 in a case, so that’s seven pounds. Okay, so they say one case. So we have to convert that to seven pounds in your head to put it into QuickBooks. And then you also had to do things like memorize the whole cheese hierarchy. So like C:J for Jack, M for Monterey, 8 for eight ounce. That’s how you punch it into QuickBooks. And yeah, you basically had to have a computer in your head just to enter an order into the computer, and it was super painful.
So we couldn’t get orders into the system quickly, and we can’t just… We’d have to highly train people and it’s highly error-prone to put orders in and people would call on the phone or you’d get a written order, and then you’d have to convert that in QuickBooks. So I said, what we need to do is just bury QuickBooks on the backend. Let’s build a really nice, basically an order entry screen in Salesforce. So we built it exactly like we want it for our business in Salesforce. So if customers want to order by the pound, by the piece, by the case, Salesforce does all the conversions, ultimately converts it back into pounds in QuickBooks, what’s integrated with QuickBooks.
But it lets you enter it in a very human friendly way, very nice auto complete. I mean, you can basically just start to type one letter of Y for yellow cheddar, and yellow cheddar just pops up. The orders almost right themselves now, and it’s very easy to train new people how to put orders in the system. So that was a huge-
Mike:
Oh, that’s good.
Daniel Peter:
Yeah. That was a huge win.
Mike:
Did you also stand up an external facing way to take orders too, to speed that up, or is that on the roadmap?
Daniel Peter:
That’s still on the roadmap. We do have, so 99% of our business comes from wholesale to stores.
Mike:
Oh, okay.
Daniel Peter:
We do have a small segment of direct-to-consumer, so you can go to springhillcheese.com and order online. Right now that’s just a Squarespace e-commerce website, and that works well, but we want to do some fancier things. So we might look into some Salesforce offerings. There’s obviously things like Commerce Cloud, but there’s also some ISVs have built some apps on Salesforce to sell. So we’ll look into some of those. Right now during this relaunch and focusing on staying alive here in the short term, really wholesale is what we’re focusing on. But there’s a big opportunity in direct-to-consumer. We ship cheese in a kind of an insulated cooler. Believe it or not, it actually ships really well throughout the US, and a lot of people do order. That’s on the roadmap, actually. But the tricky thing is the stores might not necessarily… The stores, a lot of them are still old school and the buyers still kind of want to call you. So even if we did give them a way to enter their own orders, not everybody would adopt it.
Mike:
Yeah, they’d still call you.
Daniel Peter:
Yeah, there’s still a human element there that we need to preserve. But obviously you can start to dream about how AI can listen in on your conversation and create the order for you as you’re talking, things like that.
Mike:
Quiet, AI is listening right now, Daniel, it’s taking my cheese order. And I mean, that completely makes sense. You want, for a business, you want to send truckloads of cheese, not shoe boxes of cheese. Because even a consumer, somebody like me thinks they’re ordering a lot of cheese. Well, compared to a grocery store or a supermarket chain that’s ordering for five stores to stock 10 coolers, focusing on wholesale totally makes sense. So keeping that human element there, you can always go to the consumer, but then you’re shipping probably smaller chunks of items. So you really need the volume first.
Daniel Peter:
Yeah. If people want to order just like a piece of cheese, they end up paying more for the shipping than the cheese costs. But we do free shipping over $100. That seems to be a good sweet spot. If people want to order, then it’s like say it costs $100 worth of cheese and we pay the $30 shipping, that kind of thing. So there is a market for the direct-to-consumer when you want to purchase a little bit bigger order, it kind of makes more sense. But yeah, this place can make 140,000 pounds of cheese a day, and that’s one of its-
Mike:
Wow.
Daniel Peter:
… images we need to leverage. So yeah, definitely moving by the palette or by the semi-truck load is how we can make money here.
Mike:
Okay. So you said something, and we talked at Dreamforce that I think a lot of small or medium-sized businesses are faced with, which is they all, “I’m too small for AI.” Like AI is really for the big players like the Fortune 500 or the Fortune 100 companies need AI. But you’re just a creamery, not just, I mean, you’re a small creamery out in California, and before I pressed record, I mean you were at Dreamforce, you saw all the Agentforce stuff, you saw where AI is going. What are your plans for AI?
Daniel Peter:
Yeah, I think the story we hear a lot is about AI replacing jobs. There’s a lot of fear I guess, in the job market, both on the developer side, admin, but also basically any kind of a white collar worker that might be doing any type of work is my job going to be replaced by AI? But I think there’s all these small businesses that really have a chance to have a renaissance because of AI. So instead of losing jobs, it’s more like AI can enable them to save themselves, to save an industry or to save a small-medium business. We’ve seen, unfortunately, a lot of… Fortunately, or unfortunately, I guess there’s been a lot of disruption over decades. A lot of mom and pop stores can’t afford to be around because of places like Amazon. It’s kind of changed the world, but Amazon’s a great service too.
I mean, you can get anything you want in a couple of days, so you could say it’s fortunate or unfortunate. But I think people like to have some of these old school industries around. It kind of makes life more interesting to be able to go visit a historic place instead of everything just coming from a centralized factory. Yeah, I think the story for these smaller industries is, hey, you can actually compete. AI levels the playing field a bit now. You can actually afford AI, and it’s pretty easy to use with tools like Salesforce. So you can actually make your business reinvent your business model, to be efficient enough to be profitable, this new paradigm. So that’s the story we have here, and that’s why it’s exciting to me here.
And so I think, yeah, small business, you need a little bit of tech talent. You might have to find, like my cousin that owns this creamery came and found me to help them out with that. You might have to find somebody to help you, but it’s really easier now than it’s ever been. And I think people are pretty comfortable now doing things like writing customer service emails with ChatGPT that we’ve achieved a certain threshold where the fear is kind of going away and the trust is starting to come into play. And so it’s just a matter of, okay, well, how can I implement this now at scale for my business, to where I’m focusing on the things I’m really good at? Like my vision for the business and interacting with the customers that come in the door while AI is doing all the grunt work in the background for you, and at a pretty low cost.
Mike:
So you envision at your creamery using AI as kind of a scalable factor? Like you can scale one person, as opposed to how our conversation began, which is you would just throw people at it. AI kind of gives you that ability to throw people at it, except you’re just throwing an agent at it, right?
Daniel Peter:
Yeah. It’s sort of what the cloud did for computing when we needed elasticity. When you got a Black Friday and you got a million requests all at once and your server went down, and the cloud, kind of solve that. I think AI solves the people elasticity problem. Certain times you have a lot of orders coming in, and do you really want to have 10 people on staff all the time when they’re mostly sitting around? But then once in a while, all 10 phones are ringing. AI can solve that pretty easily.
Mike:
So you saw a lot at Dreamforce. I mean, we’ve rolled out so many things, it’s hard to remember everything, but I mean, even my team is sitting back trying to build examples for agents and prompting. As a small business. And I think it’s fair to say that where are you looking at deploying, or hey, this could be a good use case for Agentforce, or really thinking about how we can scale there.
Daniel Peter:
Yeah, we’ve got a few great ones. I think they’re all similar level of importance. So this will be in no particular order, but one of them is an agent to take orders. It’s very smart. So it would be something we’d use internally. Maybe even a delivery driver in the field might dictate it from their phone, or we could use it right in the office, but instead of having to type things into the order entry screen, we can just talk to it, talk to the agent. Ideally even with our voice and just say, “Hey, for Safeway, number 572 in San Francisco, they called in, they want to duplicate their order, but they want to add in an extra case of this.”
And it has all that context already of what they ordered before. So you don’t have to go look it up and try to clone it and add to it, and it can just crank out that order. And it can say, is this what you meant? And you can say, yeah, so there’s that one we want to build. There’s also one that’s more customer focused, and that would be, we get a lot of calls, emails for random customer service requests that are… They’re along the lines of refunds, credits, changes, questions, things like this. And again, we have all that context. We had QuickBooks for 20 years, so we have 20 years of their order history imported into Salesforce.
Mike:
Wow. Wow. That’s huge.
Daniel Peter:
Yeah. So we don’t have a data problem. Actually we have plenty of good data here, but these things eat up a lot of time. So people call in and they’ll send a very vague request. They’ll say something like, we need three replacement pieces of yellow cheddar. And so we make so many different types of yellow cheddar. And then once you figure out what they need, how do you handle that? Do you add them to their next order? Do you create a credit memo? These kinds of things. So you can imagine an agent knows everything about the customer’s history. It fills in the blanks on these vague requests, and it does the action. Whether it’s adding something onto the order, issuing a credit, that kind of thing.
And if it needs to, let’s say emails the channel that could clarify through an email. Maybe if it’s a vague request and we have context, but the responses are still ambiguous, there may be one thing we need to clarify. It can do that agentic loop and iterate until the customer’s happy with the outcome. So that’s another one. And then the final one that I was thinking would be really fun to deploy, it’s kind of a… I haven’t thought of a name for it yet, but basically be kind of a cheese agent that really knows everything about all of our products. So we’ve got all the order history, but what we also have is a super rich library of unstructured data here.
So there was a world-class lab here on site for 20 years. It’s one of the three blocks that this place is on, and it had all kinds of testing equipment. And they cranked out so many analysis, studies of all the products. So anything, any random long tail questions somebody might want to ask about dairy products, cheese, butter, powdered products, any of that, we have the answer to sitting in these unstructured documents. Anything from PowerPoint presentations to Excel to PDFs, and I can just imagine putting that into data cloud and being able to produce a really good answer.
So I think probably a way better answer than a customer service rep is ever going to be able to come up with, because how are they going to be able to do all that research?
Mike:
Yeah. It’d be just based on their knowledge. And then it’s a lottery based on whoever’s working that day.
Daniel Peter:
Yeah. So if people want to know salt content and this or that, or do you use this type of rennet that had this particular genetically modified organism or not? Or how many days out of the year are your cows eating grass? Or how long is this product good in our refrigerator at 38 degrees? We can answer just infinite number of these long tail questions. So I think that one would be not just an efficiency gainer, but actually would increase our customer satisfaction by giving them not only the information they want, but these generative AI responses have a very nice friendly tone to them also. So you’d get a great answer and it would also make you feel good, because it’s so nice to you. And so I think that would be huge for customer satisfaction.
Mike:
Yeah, I mean, I was even thinking the consumer route, once you go that way is a lot of people, when you go to the butcher, it’s the same way too. You have a general idea of what you could do with this, but especially with cheeses, the right or wrong cheese can make or break a dish. And even combinations of cheese. I like watching the Food Network, and they’ll make a macaroni and cheese, and it won’t just be one cheese, it’ll be five. And it’s funny because my mom will be like, “How do they know those five tastes good together?” I’m like, “Well, mom, they’re trained chefs.” They’re not like you and me where it’s trial and error. Let’s add this in, and it totally makes it a mess. But you could even be that same way too. Oh, if you’re ordering these three cheeses, you could order these other two. And they pair really well, all five of them together. And then they’re ordering more cheese.
Daniel Peter:
Yeah, totally. And yeah, I’ve experienced that. So we have this four-way shred blend we do for the Kings, Sacramento Kings Stadium.
Mike:
Oh, wow.
Daniel Peter:
And I don’t know what it is about that mix, but those four cheeses together, they synergize. And I go out and sample it when they’re making it, and it’s way better than… You could have one, two, or three of the cheeses. But when you put all four together, it’s like something magic happens with the combination-
Mike:
[inaudible 00:26:26], yeah.
Daniel Peter:
… cheese and some spicy cheese, and maybe there’s some mozzarella in there or something. But the way they… Yeah. So there’s definitely something to getting that magic mix right. And I think that, yeah, that’s something really fun, that’d be a fun AI problem to solve.
Mike:
That would also, you think about that. It would help educate your consumer, but it also helps sell more cheese. The good salesperson doesn’t just sell what the customer wants. It also sells what the customer doesn’t know they want.
Daniel Peter:
Yeah.
Mike:
Right? Like I know I’m going to come to you and just buy some cheddar or some Parmesan, but I’m buying these two. Did you know if you mix this with it, or here’s five other recipes you could use it for. So it’s always kind of… And again, these are all things a small little creamery is here’s how we can add scalable people on with agents that more than offset the cost of it. And you’re not a thousand person company or a 10,000 person company that’s doing it, right? But you can act like you have a thousand people.
Daniel Peter:
Yeah. So a lot of the stuff you’re mentioning is things that really good salespeople are good at, and really good salespeople are hard to find. And we have some here, but yeah, they don’t scale. So yeah, that upsell opportunity, I mean people… That actually it’s a win-win. We make more money and they actually walk away feeling like they had a way better experience. They got to try this new thing they wouldn’t have tried otherwise, something like that. So yeah, that’s pretty exciting when you can have AI fill in for that lack of top sales talent like that.
Mike:
So I guess we’ll end on a fun note. What is something that you saw at Dreamforce that you can’t wait to get your hands on?
Daniel Peter:
Oh boy. Well, I know this is the admin podcast, but-
Mike:
That’s okay. You can talk about any of our products.
Daniel Peter:
I’m pretty excited about the vibe coding. Basically, I’ve been a big fan of generative AI just as an easy button for everything and seeing how good it can write your code for you now. I was a little skeptical at first just because I’ve been a developer for so long, but when I saw it in action, I realized that’s going to be the next game changer for developers. So I actually, I haven’t been using it much, and I want to start using it now, and it’ll actually help to build some of these agents that I’m talking about more quickly. So that’s huge. I mean, I’m wondering, we’ve seen generative AI disrupt so many different tools that we use.
It’s what people use now for, we saw the students using it to write their term papers for them, and it’s just gone the gamut of things. But now actually it’s smart enough to be able to write code in quality, and it’s a really high-level skill that’s going to be a huge time saver for developers.
Mike:
Yeah, I think, not to belabor the point, and I certainly have never written code, there’s always, oh, it’s going to take our jobs. Well, technology’s evolved since forever, but it’s always taken the mundane parts of our job. And that’s how I see a lot of… Like I sat in the developer keynote and watched it. To me coding, I think they demoed building a menu, I don’t know, Daniel, you tell me. If you had to sit down and code something, I have to believe that coding the menu, the mundane part wasn’t the part you were most excited to work on.
Daniel Peter:
No.
Mike:
But if AI can do it and then it can be like, okay, so here’s the 10% I can’t do, that’s probably the 10% you really wanted to work on anyway.
Daniel Peter:
It’s so true. And that’s actually been what Salesforce has been good at since the very beginning. You go back to when you would create a custom field, you wouldn’t have to go put it in the database and put it in each TML page. It just showed up everywhere automatically. And that’s all you wanted was a custom field. You didn’t want to have to write all the tedious code to wire it from the bottom to the top of the stack. So yeah, it’s just the next logical evolution of Salesforce making it to where you just get to do what you want to do and it handles the mundane things.
Mike:
Yeah.
Daniel Peter:
Yeah. Salesforce was an early adopter to that lower code way of doing things. And it’s exciting that now it’s come all the way to actually pro code can be written with your voice.
Mike:
I mean, there’s always a need for both. I think you’re kind of a car guy. I saw a stat that like 94% of the automobiles produced are automatic transmissions now. Well, that’s still 6% that aren’t, and is that good or bad? Well, I mean, some people that really like three pedals, oh, that’s horrible, why are they making all these automatic cars? They’re taking the fun away and this. Yeah, I mean, to be fair, you drive a 50-year-old manual clutch car, it’s a lot different than when you get in your car. And I mean, some of them don’t even have a gear shift now, some of them, you just press a button to start and press a button to put it in gear.
But then also the experience of driving. You can focus on the things that really help you prevent getting in a car crash, as opposed to just making the car go the road. And as a car guy, it’s two very different things. One’s not better than the other, it’s just less weight on your mind.
Daniel Peter:
Yeah. Yeah, that’s a cool analogy. I like that.
Mike:
So I mean, they’ll probably always still make manual transmission cars. If not, I’ll just keep mine going. And that doesn’t make them any less, it just means that there’s less tediousness for you to do to make the car go down the road so you can focus on some of the other things. Perspective, right?
Daniel Peter:
Yeah, no, I like that perspective on it. It makes a lot of sense. So yeah, don’t fear the loss of having to figure out how to use your clutch and your gear shift, just embrace it.
Mike:
I mean, I still think everybody should learn how to drive a manual transmission car. It does make you a better driver. I guess the opposite of that is I imagine learning how to code would make you a better admin too. So the reverse of everything has to hold true.
Daniel Peter:
Yeah. I know it gets into a pretty deep discussion about maybe how long until code totally goes away, at least from our perspectives, creating business software on Salesforce. I think I could see that even happening. And then it’s like, well, you don’t even have to drive the automatic, it’s more like the Waymo or something.
Mike:
Even then somebody has to write the software that makes it so that you don’t have to write code.
Daniel Peter:
Yeah, there’ll always be at some level. Yeah, I think there’ll always have to be. It just keeps moving further away from the people that are familiar with the business and more down to that platform level.
Mike:
Yeah. I guess I look at it much in the same of an engineering. The people that engineer and make automatic transmissions, I mean, they make automatic transmissions that can hold thousands of horsepower. And these engineers are incredibly intelligent. That doesn’t mean that the mechanic is any less of an engineer. It just means that the engineering skill moved in a different part of the business. Much like the coding skill or the non-coding skill.
Daniel Peter:
Yeah. And I think that’s good. I mean, I think people are scared of it, but I think businesses should be in business to focus more on how they can basically make their customers happier. Not so much in having to do all the tedious behind the scenes things to make it happen.
Mike:
Right. Well, Daniel, we’ve covered a lot today. Cheese, manual transmissions, business process, AI. I think if I had a wheel, we would’ve spun it and hit every corner of the wheel.
Daniel Peter:
Yeah. Well, I’m excited to keep building out these agents we’re talking about. And I’ll keep everybody in the ecosystem updated through different means. I’ve been doing some different articles and things like that, posting on LinkedIn, and so excited to get more and more of these cheese agents rolled out, and get us scaled up without having to eat up all of our margins and then some.
Mike:
Oh, look at you. Eat up all of your margins. I see where you’re going with that. Good job. Well, Daniel, I look forward to, we got TDX coming up in April of next year. Hopefully we can see one of your cheese agents at TDX.
Daniel Peter:
Yeah, definitely. Yeah, I kind of wish that I had more to show off a Dreamforce this year, but I’ve been wearing a lot of hats here as CTO of a creamery, so we need to get this stuff rolled out like yesterday.
Mike:
Yeah. Well, I mean, the cheese has got to make the money before you can buy more software. So I think you got your priorities straight.
Daniel Peter:
Yeah.
Mike:
Yeah. And you make good cheese too.
Daniel Peter:
Oh, thank you. Yeah, I love it. It’s a little dangerous working here.
Mike:
I could only imagine.
Daniel Peter:
All the cheese you can eat.
Mike:
Can only imagine. Well, Daniel, thanks so much for coming on the podcast, and I look forward to seeing your cheese agents in the future.
Daniel Peter:
Thanks for having me, Mike. This was a lot of fun.
Mike:
Well, that was cheddar than expected. So big thanks to Daniel Peter for cracking open the cheese vault and sharing how Salesforce is helping Petaluma Creamery brie-come more efficient, more scalable, and still delicious. Yes.
Remember, digital transformation, including Agentforce, isn’t just for the big cheeses, even small creameries can scale with AI and Salesforce. No provolone required. So spread the word, and until next time, we’ll see you in the cloud.
The post Why Agentforce Is a Game Changer for Small Businesses appeared first on Salesforce Admins.
More shows like The Salesforce Admins Podcast
View all
Money For the Rest of Us
1,359 Listeners
Masters in Business
2,168 Listeners
The Pitch
1,484 Listeners
Masters of Scale
3,992 Listeners
Choiceology with Katy Milkman
1,447 Listeners
Tech Brew Ride Home
960 Listeners
Schwab Market Update Audio
348 Listeners
Financial Decoder
299 Listeners
TechCrunch Daily Crunch
41 Listeners
MIT Technology Review Narrated
258 Listeners
Hard Fork
5,473 Listeners
The AI Daily Brief: Artificial Intelligence News and Analysis
591 Listeners
The Hiring Edge
25 Listeners
Tech Disruptors
15 Listeners
The JAM pod
4 Listeners