At Constellation Connected Enterprise 2022, I posted a story I wasn't expecting: a digital transformation story at enterprise scale (CCE 2022 highlight - is your organization digitally fit? Ingram Micro reveals its keys to large scale transformation).
It all goes back to a podcast grilling, where I was on the receiving end of a jugular question:
'Give me one example of a retail company that's had a successful large-scale transformation.'
I came up with only one outside possibility. But even that one, Target, has stumbled plenty since. As I wrote:
Those who advocate large-scale transformation must concede: legacy architectures (and attitudes) are a tough slog. Who has years for results these days?
But now I have a vivid use case to kick tires on: Ingram Micro. Not because their transformation is complete - but because of how they got digital traction, and got it fast.
Behind Xvantage - Ingram Micro's AI-driven platform
The key to Ingram Micro's transformation is Xvantage - Ingram Micro's platform (and data mesh). Why is this such a big deal?
- They built this "AI-driven, self-learning, ecosystem platform" themselves.
- They built it as a cloud layer on top of their legacy architecture - without any rip and replace.
- They did it fast - as in five months fast.
- It had immediate, transformative impact.
- The Xvantage platform provides the fuel for the rest of their cloud migrations - a multi-year process that can now be paced, because the platform brings the immediate value, not just internally, but to Ingram Micro's external partners.
Last year, Sahoo shared other notable concepts, such as his take on digital fitness: the need for companies to take an almost-continual digital health check. And DigiOps - the operationalization of transformation as an ongoing concern, a daily discipline that is never done. Oh, and, another keeper: "don't automate junk" - a gut check that gets more important as we advance into AI-driven processes.
As of last year, 10,000+ Ingram Micro partners were already on Xvantage - with plenty to more to go. So, I asked Sahoo - what's the progress?
Now we have more than 100,000... We just recently announced the beta of the mobile app. So we have the platform now live in 14 countries across the globe, where 100% of customers have access to the platform.
This platform push accelerated the machine learning/AI curve:
I think we have also solved some of the complex challenges in our industry, like using AI for catalog ingestion - to really have recommendations and insights pushed to our partners. That means solving complex problems, like special pricing.
Build a platform? Fine, but that bar isn't high enough anymore. The platform is the business now. Sahoo:
There is a difference between a platform supporting the business and a platform becoming the business. We are getting closer, slowly but steadily, to becoming a platform business.
An important distinction. So how will Sahoo know when Ingram Micro has truly crossed over?
The way we know crossover is when the strategy of every business owner is all tied to the transformation levers of our Xvantage platform. Every single thing - from how we grow, how we improve our margins, how we operationalize the business - the strategy is all tied to the digital transformation of Xvantage.
From day one, the focus was on external engagement:
We were really starting with the customer or partner, and then reverse engineering the platform - and we have such tight collaboration with operations. There isn't a single conversation in the reviews that we have where Xvantage doesn't come up. We are hearing more and more from our partners and our vendors - in almost every discussion - about Xvantage. That's when we know we are moving the tide and becoming a platform business. We still have a lot of work to do; it's a journey.
Measuring progress - embedding metrics for a gut check
The initial goal of partner ramp-up made sense. But now Ingram Micro is at the point where progress must be measured in other ways. Metrics that are embraced by business leaders are essential - but what you choose to measure is crucial also. How does Sahoo's team approach that? Sahoo says they have a "checklist" approach - and revisit metrics on a weekly basis:
Let's say we can just do coding, or we can just do pricing, or we can actually do recommendations. All that is okay. But is it truly showing up in the value it generates? That's very important. Ultimately, it has to be measured with value - otherwise, it's pointless. So the way we do it is: we invest in a lot of financial frameworks and modeling that analyze the data, to really give us indicators about what is working, what is not working.
How much time are the partners spending in the platform? What are they using versus not using? And most importantly, we have a weekly call, where me and my CEO, both of us - we talk to all the chief country executives and presidents about what is working, what is not working, listening to feedback - and bringing that feedback onto the product development. That happens on a weekly basis.
Ultimately, platform partners and customers determine that value. Sahoo says that the Xvantage platform has helped partners "take costs out of their business." Examples please?
Number one: creating a single pane of glass for both transaction and interaction... It actually personalizes the experience, based on the persona, even within the customer. Let's say I'm from accounting, so I'm looking at finance around procurement. The platform will be can be customized differently for every single person, or for the customer. It's a completely personalized experience.
Second, it actually takes complexity out things like conflict order or creates a single code to order in this industry. We have to process this in days or weeks. Right now we can process it in seconds, and that actually creates a much better experience. We used to get 2.4 million calls a year in US Excel for where is my order. Today, all those orders are coming from the platform, where you can track your delivery looking at the map, looking at the listing tracking number - so you can really do that on the platform itself. You don't have to call anymore.
The wrap - generative AI disruptions ahead?
Perhaps the biggest change from last year's talk? The generative AI cultural phenomenon. But as we hashed out at CCE, the transition to enterprise impact isn't as simple: CCE 2023 - the great AI debate shifts from content creation to...data governance? What does gen AI mean for Sahoo's roadmap? This is still playing out, but he does see immediate potential for enterprise search:
We are using some generative AI in real use cases. Let me give an example: search in our platform. When we search, we have to improve relevancy. We have so many SKUs and items, we need to bring the right results, personalized for the customer in the front page - like the top 10 results. But if you look at it geographically, the data is all scattered and not right. So we are using our own LLM models to really write and improve the relevancy, and bring those results back to search - so that the content that we create from search is much more relevant for our customers.
And yes, that means data cleansing is now a top job also:
We're really working on such use cases, focused on our real-time data mesh that we've built to clean and have good data. We are evaluating other use cases with gen AI as well, but this is one use case we are already working on.
Search is not typically my favorite gen AI use case, but Sahoo's focus on data quality - and allowing users to upvote and downvote output - could make for a solid result. There is more to say on metrics; Sahoo says their real-time analytics allow views into retention rates, average order value, and super-granular items such as increasing basket size and activating dormant customers. Relevant analytics are also presented to customers for their own self-service inquiries.
During our talk, Sahoo made distinctions between platform-only, platform-also, and platform-first. Given Ingram Micro's existing footprint, platform-only was never an option. Sahoo's team is building momentum to make that move from platform-also to platform-first ("It's a mindset shift," says Sahoo).
But if you want to become a platform company, you better pile up proof points along the way. As Sahoo told me:
We launched the vendor platform this year as well. Our vendors, like the big technology brands, they are logging in, and looking at performance metrics. They're looking at how their marketing campaigns are working, looking at their analytics. So that has been really successful.
So far, so good. I'll look for the update in 2024 also.