This week, we're going to be talking again about that little devil that keeps cropping up - data. 

I'm speaking to Costas Xyloyiannis, CEO of HICX, a company that takes a slightly different approach to data. Today we’re talking about what he thinks is the safest way to get clean data in your organisation.

Before we dive into the specifics of what makes his approach to data so different, I start off by asking Costas to explain Hicx in a nutshell.

Costas explains that HICX is in the business of supplier experience management. Supplier experience, for Costas, equals data. The end state isn’t just that data - there are value drivers after that too - but data is the foundation of supplier experience management: Better data means better experience, and a better experience in turn yields better data. It’s a flywheel value effect for both sides.

I ask how this differs from other solutions that take the approach of automatically gathering data using scraping techniques or AI. 

Costas gives a few differences here. Number one, what is the customer trying to solve? A lot of his customers want the right data in their systems. What you tend to find is that when you pull data out, clean it, change it, what you find is that you can never put it back into those systems. The data will have changed. So this is a very high risk, unsustainable approach. But this is the way most people have done it traditionally.

HICX puts processes in place which control how data is entered into those systems. The supplier is the source of truth, so why not optimise the process of collecting data from suppliers? Only then does HICX apply automation to enrich that data.

I ask Costas to clarify that he is using some degree of automation, but the fundamental difference is that he’s relying on the supplier to provide the core master data.

He says this is correct. He then goes on to speak to customer demand. Customers need a very granular view of their suppliers. If you’re looking at, for example, a process around manufacturing that has to take place at your supplier’s facility, most other sources don’t explain the things customers want to know: What is the parent legal entity, for example? These factors change how data and business processes are managed.

I bring up vendor master data. We often think about this wrongly: “What data do we need in the system to pay the supplier?”. But in something like manufacturing, food or automotive where you’ve got health and safety requirements that are important to the qualification process, there has to be some way of distinguishing the supplier experience and on-boarding process. If you’re using a tool that automatically cleanses data, it’s not going to know how strategically critical a supplier will be to you.

Costas agrees. He also says that context of how you use data is important. It could be the address of a supplier, it could be a payment address… When you’re cleansing you don’t know these things. Costas thinks that this is where customers need help: Who is the parent entity? Is this supplier part of the same legal entity? What is this address? How does it all fit together? This is what customers need to understand to a high degree of accuracy.

He goes on to talk about the importance of being clear on your use case. If your use case is highly analytical, then using an outside source to cleanse your data makes sense, because your goal is to process a lot of data. If you actually want to change the data in your operational systems, this does not work. You need context to meaningfully make those changes. What does each data point mean to your organisation, and depending on context.

I mention HICX’s clients, who are largely enterprise-level organisations. If the suppliers are HICX’s source of truth, then what typically does Costas see from his customers when it comes to managing or changing things? If you have too many people involved in this process, do you get to a point where you’ve got incorrect entries? Is there a master data steward approach, or a centralised department in charge of what data gets entered into HICX? 

Costas says this is a great question. And that the key thing is that suppliers are very cross-functional. This is why supplier master data is harder than customer master data. Everyone touches suppliers in one way or another. Around data governance and ownership, you need to have a global centralised team - that’s just the efficient way to do it. How to see it is that the supplier owns the data. You’re just validating it. 

But people want a lot of other information, environmental information, or other industry-specific needs. And what we’re seeing is people complementing this with “curated” data sources. Like Ecovadis, for example, they curate data for ESG.

And this is cross-functional. This will then potentially go to a sustainability team or a health and safety team, or whatever applies for a specific industry’s needs.

So customers are bringing the two together, they’re interested in driving their own processes, they want to do things cross functionally, and they want to use these curated sources when it makes sense.

I ask if this makes HICX the single aggregator of all of that data in one single source of truth.

Costas says that it’s not just aggregation, but orchestration of the process across systems.

It’s also about how data is implemented into workflow, and how data-driven workflow can contribute to overall efficiency.

I ask if there’s a risk here that vendors might see HICX as just another platform that customers are asking them to use? Given how reliant HICX is on supplier data, how does Costas get vendors to input this master data?

He thinks it comes down to context - first of all the supplier will have to supply this data to the customer. If they don’t have to do this manually through Excel or similar, you’re making things easier for them. There’s one place to go to for all their communications. You’re helping them navigate these systems, and that kind of accessible ease of use is already of value to suppliers.

 

I agree - having a single source from which to self serve is going to be a huge benefit. But then, the additional conundrum to that is “what do you do with the long tail?”.

I ask about how this plays with a platform-centric approach. Looking back on what Dr. Elouise Epstein had to say, and her model where one platform (usually a legacy suite) acts as the center of the procurement ‘spider diagram’ - does Costas think HICX could fill that role as a single source of truth that best-of-breeds then plug into?

Costas says absolutely. He thinks Dr. Epstein would absolutely put HICX right there in the middle of that diagram. HICX connects the common thread between best-of-breeds, and differing company needs, which is the need to orchestrate processes. This relies on data in order to be efficiently automated.  Costas says that suites will not have the right information to do this. 

He says that people forget where these suites and ERPs came from - these are largely transactional systems, used for orders, invoices and payments. So these systems focus on optimising the transaction. But where we are now, we’ve moved beyond the transactional. You need the right hierarchy of data and information, and the suite doesn’t deal with, for example, a parent entity or a manufacturing entity, or any of the things we were just talking about. 

I agree - if you put garbage data into a system, you’re going to get garbage out. It’s that nightmare scenario where you pay to run spend analytics and the biggest slice of the pie chart is “uncategorised”. We’ve all been there at some point in our careers!

 

We take a quick ad break, and once we’re back, I ask Costas if it’s lack of resources that’s part of the problem, or a lack of prioritisation when it comes to data?

Costas says that yes, historically procurement has not seen data as a priority. Focus on measuring success through savings means that data isn’t addressed until the problem is out of hand. Data solutions are no longer a nice-to-have, and now we’re seeing it prioritised more. But for years, bad data has been the main barrier to digitalisation in procurement. The penny has dropped, and the solutions on the market now are making it more feasible to invest in data.

I ask if Costas still sees a reluctance to invest in data. CFOs tend not to think long term, especially if they’re having to report results by the year or even the quarter. If this short term thinking is so prevalent, and the ROI on data investment is so longterm, is the worm starting to turn?

Costas says that there are savings to be gained from proper data management, and especially when you make data gathering a part of your on-boarding process. When you put it in that context, there are huge savings to be made when it comes to efficiency gain. It’s about how progressively organisations are willing to think. In some of Costas’ use cases, he’s seen companies remove seventeen full time equivalents in data management. 

 

I agree that being able to move people onto more value added tasks definitely generates savings. And before closing out the podcast, I ask Costas about the difference between enterprise-scale data problems, and the data problems that could face a small to medium sized business.

Costas says that, when a company is small, chances are that their challenges will be small. But as volumes and systems grow, you need more information, more analytics, and new kinds of insights. When you’re small, you might have enough data to derive useful insights, but you still need orchestration, especially if you’re going to grow in future. 

Even small companies should start with discipline around data, so that problems don’t sneak up on as a sudden and huge issue down the line. If you don’t act, these things can snowball quickly.

I finish off the podcast, not just by asking where you can find Costas, but also with a cheeky plug for his own podcast, where I myself have been a guest. 

You can find all the relevant links below, so be sure to check it out!