Data Syndication and Standardization in B2B: How Not To Waste Time Chasing Red Herrings
The B2B eCommerce Podcast
Data Syndication and Standardization in B2B: How Not To Waste Time Chasing Red Herrings
Yoav Kutner: Welcome everyone. Just to recap for some of the new people who are joining us why we decided to do this podcast. B2B UnCut is something that will serve the B2B ecosystem. We feel that there’s a lot of noise and sensationalism around B2B commerce and digital transformation. And because we come from this field, we hear these questions from a lot of our customers, partners, and developers that we work with. We just wanted to make a place where we can voice these concerns, talk about them openly, and have a discussion about these matters that we all are facing on the day to day basis while we work in B2B. We’d love for the audience to participate. Whoever joins us on Zoom can actually participate in the discussion.
So, for today’s topic, we’re going to be discussing data in B2B commerce, something that a lot of companies are struggling with as we are learning when we’re working with them. There are a lot of systems involved in the production, packing and storing, shipping, warehouse management, order management, etc. As we all know, they produce a lot of data. We need to have data that talks between those systems and is updated across the systems. That’s something that we want to cover today and figure out how this works, how to approach these types of integrations and make sure that your business is actually ticking well with all the systems that are involved today.
We have two people that we actually work with Oro and our first guest is Joe Albrecht. Joe, can you introduce yourself?
Joe Albrecht: Yeah, certainly. Thank you, Yoav. My name is Joel Albrecht, managing partner and CEO of Xngage. We are B2B eCommerce implementation services.
Yoav Kutner: Thank you, Joe. And our second guest from Akeneo is Ali Hanyaloglu. Please introduce yourself.
Ali Hanyaloglu: Absolutely. I’m Ali Hanyaloglu. I am Senior Director of Product Marketing at Akeneo. I’m based actually over the Boston office. But I’m originally from London, England, as you can tell by my accent. I’ve been in the commerce and SAS technology space for quite some time, and at many different companies. It’s a pleasure to be here. Thank you all for having us.
Yoav Kutner: Thank you. Let’s jump in. So the first topic that we wanted to cover is, why do data syndication and standardization matter? I think that’s something that is interesting for a lot of people. We definitely have the two companies here that have a lot to do with data and integration.
Ali Hanyaloglu: So why do they matter? I think they matter because there’s something else that matters even more than that, that is impacted if you don’t pay attention to these two. And that’s the expectations that customers have of B2B organizations that they’re working with, with respect to the data about their products, their company, their brand, and so on, that they use to make decisions. Not taking into account how customers are finding out about you, getting to know what you offer, and not being consistent, or even correct about the information that you provide to them to help them make a decision to do business with you is not only just going to be a bad experience, they will actually abandon you. So that’s why these matter. Getting these things right is going to have downstream impacts on your ability to grow as a business.
Joe Albrecht: That’s a great point, Ali, and I can offer the perspective of our clients. We work with a lot of distributors. Obviously, data is key to selling, to providing great experiences for customers in B2B. It starts with research. And it continues on with decision-making, as you pointed out.
In B2B, you have complex product data. But I want to get to the heart of the question, which is about syndication and standardization. I think that’s an important piece that the industry has not yet fully captured, right? There is a lot of standardization in software development, and standards in certain industries in business. There’s ebXML, EDI, and these are all standards to basically make processes more efficient. I think there’s a lot of efficiencies to be gained in the topic of specific product data standardization, and harmonization. Because I see those struggles every time with every single client.
The question is, do we have the data? If not, where do we get it? And then when that’s answered, and we are getting it from sources, then the data comes in all formats and shapes. Some industries have already achieved standardization like automotive, where there are ACES and PIES. Other industries still lack a bit in terms of standardization. So I think the more standardization will be achieved in the verticals, the better. Our clients and hopefully some of the folks who are on the call can manage this very significant data challenge.
Ali Hanyaloglu: Yeah, Joe, you’re absolutely right. With that respect to standardization, there is standardization, if you will, across an industry or a particular segment to help those businesses be more efficient and accurate with what they’re doing. But there’s also the standardization that you do within your organization as you work with that data. So that you are more efficient when it comes to working with that information that you’re being provided in a standard format is a starting point for you. It’s the way you differentiate yourself in the market, the way that you help your buyers to make informed decisions with you and get the right products so that they’re not returning them to you.
Data standardization forms the basis for everything to work efficiently and expand on that core technical information that’s provided to you. You don’t want to be duplicating things, you don’t want to be creating redundancy. And you want to ensure that you are still applying governance in terms of the information that gets created, so that incorrect, or incomplete information doesn’t slip through. Standardization is an important thing that you do within your company as well.
Joe Albrecht: Right. The perspective I shared is probably more the one of a distributor that gets a lot of data or has to source it from different suppliers. Manufacturers have exactly what you’re saying. And I think since both are looking to do more of that standardization and synchronization, the interchange of data becomes faster, easier, and actually less costly.
Then what you’re talking about is internally having a standard approach to product data management to create a single version of the truth. That’s a nice phrase. But yeah, to enrich them and build out the content to ultimately set themselves apart. If everyone has the same data, it’s not going to be very unique to your business. So what you’re talking about is actually standardizing how you internally manage it, but then creating an enrichment that is unique for your company, or within your industry. Look what Nike is doing in terms of how they engage and standardize their content. But also they differentiate in the market.
Yoav Kutner: What systems and data sources do you see in big projects that normally you’re presenting? With a PIM solution, Joe, you probably are sitting in the middle when you’re helping your customers build these projects. What would you list as systems that you have to integrate and have data moving back and forth?
Joe Albrecht: Working with lots of clients, it depends on what type of data we’re talking about. So if we want to quickly start with product data, obviously the systems that folks are adopting to standardize are PIM systems. When it comes to sourcing that data on the distribution side, there is actually third parties in the ecosystem. Their business model is to aggregate sources and enrich product data to then provide it to the various members of the community.
For instance, one I can maybe talk about quickly is ADE Data Center. These are very important sources that we see frequently. When it comes to other data, obviously, the systems are different: you’re dealing more with ERP systems for all the customer muster, you’re dealing with CRM systems, maybe a bit order management systems, etc, as we venture away from product data.
Ali Hanyaloglu: Yep, those are two grids. That’s actually a very interesting example that you brought up, Joe, that many people don’t consider. We actually have a customer of ours who provides that exact service, which is providing product catalogs to those in the manufacturing, food service industry, and so on.
But in terms of other systems that we see, as you said, there are all different types of data you could even ask like, What do you mean by product data? There are all different types that can be considered there. So we primarily see outside of the PIM world systems like ERP, and PLM, of course, on the manufacturing side. And believe it or not, we still see spreadsheets out there. A lot of data is sitting in spreadsheets and PDF files. It’s still out there.
Joe Albrecht: We see a lot of product data being aggregated in WordPress web pages as well.
Ali Hanyaloglu: Part of the challenge that our audience has is how do you remove the silos of all those different systems of information? That’s the challenge and what needs to be addressed for us so that you can then actually standardize and, therefore, more easily syndicate the information.
Yoav Kutner: Another big question that a lot of our customers ask us, and I’m sure you’re hearing it a lot is when do we actually have to start thinking about data standardization? How do we synchronize data? How do we manage data? When is the right time to start thinking about it: when we are going into digitizing our business or going for a new project? So I know my thoughts about it. But being a moderator, I want to hear your side of the story about this and how you think about that.
Joe Albrecht: Yes, it depends a bit on when you start to ask this question. Product Data, for instance, is a critical asset to do anything online in terms of selling online or even supporting research processes online for customers. So the answer for when to start synchronizing product data is yesterday.
As a manufacturer, you probably don’t have this struggle, but so I think the answer is also yesterday. If you’re asking this more specifically in the context of a project initiative, like a Commerce project, it should be as early as possible. Product data informs a lot of the dependent processes such as the user experience. Without understanding your data, you cannot decide and design the most suitable selling page. Because if you don’t know what’s the item that you added to the cart, you’re done. Or it can be more complex, where you’re choosing based on attributes, facets, options, or maybe even configurations. If you don’t have the answers for that, then how can you optimize the experience?
That’s one big consideration why product data early on matters. The other is very technical. You need to get this product data early into the systems so that the experience can be built out and that the integrations can be built out. Of course, if you get a PIM platform right then you have a leg up there. Product data is a strong dependency on the success of the project.
Ali Hanyaloglu: I wholeheartedly agree. I know the analysts that we talked to also agree and make this recommendation to their clients that they speak to, when not to do this. It’s not something you do after you finished your replatforming, or digital transformation effort. Don’t get that technology in place, and then say, Okay, now let’s clean up the data. That’s garbage in – garbage out.
The recommendation is, do it before you make those big investments in not just the technology, but your people and the resources and the time. And it’s been proven, if you get your data sorted out first, your actual digital transformation or replatforming effort will go faster. Because everything is as expected. And as Joe correctly said, it’s not just out to your eCommerce platform and check done. There are so many other systems that will be dependent on that. The other reason to get it done is that before you fully finish out your replatforming or transformation is you’ll be ready for when this needs to happen again. It will inevitably happen again, and you’ll be ready for making that particular transformation as necessary.
Joe Albrecht: That’s exactly right. One more comment on it. This standardization effort should be ongoing. The reason for that is that the usage scenarios are changing, and customers demand increasingly different data points from the sell-side. In the past, it was great if you had a PDP and there was a picture. Today, that’s not cutting it, you have to have spec sheets, collateral, and key information that drives decision making. We need to consider the 3D world, you need tangible models and assets that go with that data. So you have to look at your product data constantly.
Yoav Kutner: Do you think it’s only limited to product data? I think this is even becoming a bigger problem when we start adding other pieces of data. I know a lot of this actually drives many business processes of a company in the end. Let’s say we convinced our audience to start thinking about data is crucial, but how do we start going about it? It seems very messy in terms of what we need to think about how much we have to consider, and how many moving parts we have. So what will be a strategy with which we can actually start this process in a company?
Joe Albrecht: It starts with the use cases and understanding what you want to achieve. If you are a distributor, selling 2 million SKUs, is your goal to manually enrich your data? Probably not. What your goal would be is to ensure that you have as much of a data set that’s complete, consistent and cleansed to support 2 million SKUs on a sell page.
Then your challenges are all about sourcing that data and standardizing that data, where we go back to the first point of the discussion here. When you identify that you need to take care of your product data, the enrichment is your use case. That’s the time when you look at PIM platforms, but you still want the use cases to guide you in that journey. Because some products are strong with enrichments. Others are strong with the syndication, others are strong with asset management. And then some are strong on the inbound side. So depending on where you are in that chain of how the product data is used, whether you’re the creator, on the manufacturing side, or the consumer on the distribution side, your use cases will be weighted differently. I think that can be a good guide to help you with the process. And then it’s not just about the technology, it’s building a team around it. I would call it a data discipline in-house, maybe a data wrangler or the team that is truly there to solve the data challenge.
Ali Hanyaloglu: I see it being on two dimensions you need to be looking at and a very similar philosophy that Joe just shared. It’s all around how do you define the strategy that you’re going to take. And that’s where you start. And I see two dimensions. One is, to think about it in terms of scope for how the data is going to be used. That could be, for example, you’re a multinational business and you sell in different countries. So is your data structure so complicated that you’re trying to account for every different language and every different nuance. How you structure your products with their overly complex attributes, and instead, looking at things like locales, whereby you only have one description, but it has different locales supporting it, that makes things more efficient for you.
Maybe it’s in terms of you actually wanting to grow by adding more channels to how you market and sell. We’re seeing this more and more in B2B. And starting to embrace some things that may have been considered B2C. But it’s not just that we’ve got an eCommerce site and then a direct Field Sales Team. They are looking at more and more and saying, Well, actually, we want to get on marketplaces, we want to sell through partnerships, we actually want to start marketing, even through social media. In a B2B context, these will become part of the scope of defining the strategy of where to start and how you’re structuring your catalog.
The other dimension is more around your goals and KPIs. So identifying upfront things like why you’re doing this. And it could be things like we just need to get products to market faster. It’s taking too long, and we’re losing business as a result. So what do we need to do not just from that data perspective, but to Joe’s perfect point, which is what people and processes that we need to be able to put into place to support getting those products to market faster. Maybe you want to sell more products, maybe you’re acquiring brands, for we see this a lot. And so you need to be able to make sure that as you acquire new brands into what you offer, the migration over to a single catalog becomes easier if that’s your goal and direction. The list goes on. But define those two things first, it will then determine the strategy you’re going to take with respect to managing, standardizing, and instantly tracing your data.
Yoav Kutner: From your experience, guys, who do you think has to own this process? What’s the successful way of having multiple owners, so that the integration can touch every part of the company? What have you seen was successful, what was not so successful in having a single owner or multiple owners, etc.?
Ali Hanyaloglu: I’ll take that one first. Remember, there are two dimensions I mentioned. So who’s going to lead it and drive it is the ones who are accountable for those two dimensions that I talked about. Whether it’s your scope, and what you’re doing there in terms of growth or your particular business objectives. That’s who could initiate or drive this particular process. In many cases, we typically see that more on the line of business or marketing side. There needs to be a partnership here between the business and the IT side in order for it to be successful. And what we typically see is that we’ll have a lead from one of those two roles, but they partner closely with their counterpart and take their needs into account in order to drive successful projects. That’s what we typically see from the PIM side of things.
Joe Albrecht: This should also be backed by leadership, who is ultimately responsible to make sure data is an important topic. Data is a currency nowadays. Whether we talk now about product data, or other data in the organization, but product data alone is an asset. It’s the IP of an organization. Without that data it’s going to be very hard to succeed in a digital world. So I think it’s a leadership topic.
When we talk specifically about synchronization and syndication, or the end center standardization, you’ll need a product data wrangler or a cross-functional Center of Excellence. Take product data on a website: if it’s just technical data, nobody will be able to buy based on it, but if it’s not technical and only marketing, it will never work for the real use cases that the customers out there have. So you have to have the rich, attribute-heavy, very detailed product data sets that typically product data managers own and maybe come out of PLM and ERP, but you also need the marketing side of it. It’s a cross-functional topic.
Yoav Kutner: When we start these processes, we start doing data integration. And there are a lot of times where we see that many of the systems that are involved actually generate the same data, right, or same data that can be regarded as duplicate. And there’s this term that comes up a lot in integration – the source of truth, meaning which system is actually responsible to make sure that a customer name is the customer name, or SKU number is an SKU number, etc. And we see many, many different systems that can be sources of truth. There’s always the question of how do we define what is the source of truth that we should refer to?
Ali Hanyaloglu: Nobody creates a source of truth or a system of record, as it’s sometimes called, for the sake of creating it. There’s no such thing. It could be in a number of places. And the way I tend to look at things from both a business and architecture perspective is how is this information going to be used. Therefore, one or the other systems that are going to be leveraging it as well as what is going to be measured in terms of its impact. Then work backward to what technology makes sense, that is going to be able to leverage that information, as seamlessly as possible without having to jump through massive hurdles.
We have customers who do use PIM as a source of truth for their product data. Interestingly enough, we also have some customers using PIM as a means to collect all the information, and rich it as necessary. But then it ultimately gets pushed back to the ERP, for example, they may still use other systems to call from the PIM but the corporate source of truth for that product data actually lives in a different technology, and it’s been pushed out and stored there. So you can see the difference, right? It does depend on how you want to best leverage that data and what makes sense for your organization.
Joe Albrecht: Yeah, I don’t know what else to add there. Other than that’s probably why we exist as a services partner. You can’t just say, Oh, I have a centralized payment, this is the source of the truth. Because the eCommerce platform comes along and can offer certain things that maybe the PIM also could do, like product relationships, merchandising associations are a great example. Are you managing your cross-sells in the PIM or in the eCommerce platform? There is a lot of overlap in some of these areas. You have to decide what’s the source of truth, and then configure and build processes around those decisions. So your data might be flowing into the PIM, that’s the centralized place to aggregate and enrich it. But then you need that data in the ERP, but the ERP cannot hold it. So you end up with a data mess on your hands. That’s exactly what the implementation goes through.
Yoav Kutner: I completely agree with both of you. The source of truth is something that’s evolving and can be in different parts of the project and in a different system. That’s definitely a complex one that a lot of people are expecting a very simple answer, and I don’t think that one exists.
Have you ever experienced projects which went wrong, where there was bad data or bad data integration? I’d love to hear your horror stories and see what has your experience taught you?
Ali Hanyaloglu: One was a case of what we saw online, whereby some dog food was being sold in both a B2B and B2C scenarios. So the product description page for this had the dog food labeled as both adult and puppy food. Now I’m no vet. And if you’re not a pet owner, maybe not aware of this, but I am a dog owner, it is very dangerous to give an adult dog puppy food. And so this was one thing that we had caught. And thankfully, it’s been corrected, of course. But this was a case whereby, because standardization, or any governance or anything like that was not being put in place, mistakes like this can slip through.
It’s more than just looking bad on the brand for not being clear with their information, or clearly potential legal implications of that particular scenario coming up. And the thing is that if it slips through, we can very easily get through to other channels without being realizing that and sometimes they can be hard to correct. So that’s one particular example I share where it seems like a small detail could have big implications.
Joe Albrecht: The tricky thing is sometimes just a little bit of bad data can cause trouble. So there are different ways to express units of measure. I’m not talking millimeters versus inches. I’m talking, are you saying “inch”, or “in.”? Not standardizing the terminology alone can cause headaches. I have seen one example where folks pump in product data and it turns out that the titles of the products had credit card numbers in them. So that’s the red flag.
And then also just bad data. Bad data is very common. By bad I mean either gaps in data or simply very abbreviated data that nobody understands other than maybe the sales rep that has sold this thing for years. So it’s all over the place. And bad data is actually the reason why there’s a ton of churn in many of the implementation efforts. And in this topic of integrating systems and having data flow, bad data is your enemy.
Ali Hanyaloglu: And then one of the biggest business impact risks that you have with bad data is an increase in returns of the product that was bought. And that is a very costly proposition. Now, remember, there was a number that said like it’s in the hundreds of millions of dollars are spent each year on just on returning stuff.
Yoav Kutner: Yes, returns and bad data can kill you. But, again, this episode is also about the positive. So I would love to hear some good stories, let’s say what would be a case where you did come through a company, and maybe from your experience, you help them succeed in data standardization.
Ali Hanyaloglu: This is a story I do like to share given the last two years. And sort of very quickly on this one. One of our customers is Socomore. If you’re not familiar with Socomore, they’re based in France. They’re in the business of providing cleaning goods and services to the airline industry. When the first lockdown happened when the pandemic hit – no airline industry. This company could have gone bankrupt overnight because of it.
They actually decided to pivot very quickly and what they did was they took their existing facilities and services and started providing sanitization gel to local hospitals and other medical facilities that needed them. So they were able to have a business there. But word got around that they were providing sanitization when it was very hard to get hold of, and they actually not only pivoted to a different industry, they actually pivoted to B2C. So they actually started selling to local stores, and then to consumers as well through their website, a complete pivot that actually helped them survive and continue to sustain their business through the pandemic.
They have credited a PIM, obviously, Akeneo PIM, in this case, in having their data in order already, they were prepared to make that pivot. Because they had all the information about their products and what they do ready to go, it was easier and faster for them to be able to pivot to a whole new business model, to a different audience, rather than having to wait to recreate all of the information all over again. So that’s one happy story that I do like to share.
Joe Albrecht: I don’t have a great story right now readily available to share but good data management is so foundational. When we are engaged with clients and we do our discovery process, we try to understand what the data situation is. And I’m now not just talking product data, but in general, and we’re looking for the quality of the data, the availability of the data. Can it be accessed? Can it be made available through APIs?
We have been fortunate enough to work with a lot of clients who put the topic data at the forefront. And again, that goes back to the earlier comment. It’s part of the leadership focus.
Yoav Kutner: All right, thank you. What are other use cases that data management is crucial for and can save a lot of time and resources? I’ll give a quick example that we learned from one of our customers, a big distributor in the US. We went to their tradeshow, and we saw that their customers were coming in ordering from books and then filling in manually orders on the order sheets. And then somebody had to key it in the ERP. It was so inefficient. What we did was create a digital funnel where all the data comes in from and then it gets distributed to the right place. I think that was the easy way. So what kind of other use cases do you see where good data synchronization and management were crucial for business?
Ali Hanyaloglu: We talked about pivoting business models. We’ve talked about replatforming. And digital transformation is another example. I gave an example whereby whether it’s a manufacturer or even a distributor is bringing on a new brand or a new line of products. Again, getting the data governance and organization as part of the management will make those particular projects and initiatives go faster, and minimize the number of resources that are needed to bring on that new brand or product line, which in and of itself can be a huge project. Those are probably the main ones that I come across.
Joe Albrecht: When companies go through digital transformation with eCommerce and they go live, there’s this expectation that now we have an eCommerce channel and the orders will fly through the door. But that’s not happening. Why? If you don’t have a data mindset and data focus to even measure how the web property is performing, you’ll have a hard time making the necessary changes and coming up with the necessary strategies to create that lift on the acquisition side. You might be still doing good and the adoption side is getting customers that you have relationships with over the years to channel shift and adopt the web platform and they become beneficiaries of your digital transformation journey. But if you want to write the acquisition story and penetrate new markets, and all this without data discipline internally, and the focus on it, this is going to be very difficult. So I’ve seen a lot of success stories where mature clients do a lot of that monitoring and measuring so they can tell what are the top-selling SKUs, what are the categories that are purchased the most frequently by certain customer segments. These are things that are important, and then not last component of that story is also SEO and analytics data.
Yoav Kutner: The big question that I guess that we’re gonna end with is, why are data and content important in merchandising in the B2B space?
Ali Hanyaloglu: Funnily enough, it’s the same reasons that apply to the B2C space as well. And it’s the same in B2B, it’s no different here. Merchandising in this case needs to be correct. First of all, we do see examples, both in B2B and B2C, where there is some evidence of merchandising going on on the eCommerce site. And it’s like, what are these products have in common? So there’s the correctness, there’s relevancy as well. Today, many of the technologies that are driving, eMerchandising, including things like searchandising, have to be relevant to those who are coming in, in order for you to grow average orders, deal sizes, and get more business.
We want to make sure that the strategies we are taking with that data are having an impact on our business. Otherwise, why we’re doing them. And the way you do that is through analytics and the insights that come back. Merchandising is a means to be able to get some of those insights and analytics come back into your business in order to determine how you refine your strategy when it comes to data and what you do with it. And so for all of those three things to be effective the data, the content, the performance data, you need to ensure that analytics is working as expected.
Joe Albrecht: What’s more important than the data itself is the relationships. And well, the relationships also depend on the data. So if you don’t have attributes, if you don’t have facets, if you don’t have spec sheets to relate products with each other, it’s going to be a problem in B2B. Business customers are not just buying a product to satisfy some personal need. They buy products that need to be fit into a process, whether it’s being manufacturing process, construction process, what have you. So they need to find the right thing. And that’s a complex discussion. That’s why sales reps exist, while you can’t put a sales rep on the website. So if you want to sell without the sales rep, being actively talking to you on the website, you have to have a lot of that data that helps your solution.
Yoav Kutner: We’ve discussed quite a lot today. I wanted to thank you both Ali and Joe, for participating in this episode. We are going to have our third episode coming up soon. So it’s going to be on B2B commerce adoption, and connecting your internal and external stakeholders.
Thank you everyone very much and see you next time on B2B UnCut.
- Why does data standardization matter? It helps businesses be more efficient and accurate with what they're doing. It’s the way you differentiate yourself in the market, the way that you help your buyers make informed decisions and buy products that they won't return.
- Part of the challenge that our audience has is how do you remove the silos of all those different systems of information? That's the challenge and what needs to be addressed for us so that you can then actually standardize and, therefore, more easily syndicate the information.
- When is the right time to start thinking about data standardization? It’s not something you do after you finished your replatforming, or digital transformation effort. Do it before you make those big investments. If you get your data sorted out first, your actual digital transformation or replatforming effort will go faster.
- There are two dimensions to data standardization: one is within the scope of how the data is going to be used. The other dimension is more around your goals and KPIs. So identifying upfront things like why you're doing this. Define those two things first, it will then determine the strategy you're going to take with respect to managing, standardizing, and tracing your data.