🇺🇸
A realistic look at the real roadblocks in a supply chain planning migration: objectives, data, trust, and planner judgment.
Transitioning from Excel to an advanced planning platform is one of the biggest hurdles supply chain teams face. And it’s not always the technical or data side that acts as the main barrier.
Among other reasons, it’s because the time required to move operations to a platform is underestimated, leading to a backlog of work. Then, the uncomfortable question arises: How can I trust this new platform when I used to have total control? When that isn't resolved, adoption fails. You might think you're transforming your operation, but you're actually just wasting time, because true transformation starts with how you work.
In this article, we reflect (having spent years doing this same work in Excel before moving to the platform side) on what we’ve learned at Datup while helping companies move from Excel to an AI-powered Supply Chain Analytics platform.
"I spend too much time in Excel processing information and want to implement an advanced supply chain analytics platform".
The problem is real, it affects many industries, and we see it every day with every new client. Operating a supply chain is complex, but fortunately, you have a wealth of data you can rely on to make effective decisions. Now, before that, what should you keep in mind to ensure a successful implementation?
This is one of the most fundamental topics. When you are clear about them, it doesn't just become another task: it becomes a final purpose.
One of the pain points our clients express most often is "I spend too much time in Excel processing information". But you need to make it more tangible:
Because when you have that North Star, the project becomes important to you. And the time you invest in it is time you invest with enthusiasm, because it will be for your own benefit.
And to avoid seeing it as just another burden, you have to be aware of this from the beginning.
Excel allows for a lot of flexibility: you can adjust, you can create a new formula. In that sense, it has potential. But there is also too much manual work when dealing with repetitive tasks and large volumes of data. It slows down operations and delays decision-making.
If you want to explore data in Excel on your own, I think that's interesting, because you can reach some cool conclusions. But when it's a repetitive task that you always have to do:
When you put it into perspective, it's a lot of day-to-day tasks. If I don't make time for the new project, I'm going to stay stuck in this, or I'm going to drag both things out.
You already know what you're going to achieve and how it will impact you: in terms of manual work, efficiency, and human error. And we never really measure that last one: the decisions I got wrong that had an impact, simply because we spend our lives putting out fires.
Do you want to keep putting out fires, or do you want to move on to having a tool that helps you in your day-to-day?
I'm talking about 15 years ago, when I used to do that in Excel. Back then, I would have loved a tool like this. But today, adapting to new technologies and getting the hang of tools like Datup is just as beneficial.
“We have our goals clear and we are going to implement the software.”
And that is where you run into a major barrier: data connectivity.
There is a business and process aspect to adapting to new platforms, but many ignore the technology side. When you integrate with a platform, it's not just your business side that's involved: there is a technological data issue behind it. And when a decision-maker decides to "move to an advanced tool," it's not the first thing they take into consideration.
That is where many companies get stuck: there is no data fluidity, and they hadn't placed importance on having a clean database to be able to integrate with these new tools.
Implementation requires two profiles:
Often, they don't even want to take ownership of the process. "You do it,"they've told us when it's time to implement the demand planning module. But without proper documentation of the data structures, the process drags on, and a configuration that could be relatively simple ends up taking much longer.
The difference is quickly apparent.
We've seen it from experience:
"I have the process perfectly clear; I've already built it in Excel. I just need to move it exactly as it is to the platform."
If you want to literally duplicate your Excel process on a platform, you will likely run into certain limitations.
We usually guide our clients in defining their forecasts: "Look, how do you plan? What are your locations? What are your SKUs?". But if planners don't focus on fully understanding or breaking down their demand planning process, that also creates bottlenecks and delays, making the data delivery process much more drawn out.
It is also very important to know what you are looking for. Software as a service offers benefits such as speed of setup, but you might not be able to adapt it exactly to your business. You can also leverage the software's experience (like Datup), which often understands multiple supply chain operations across various industries, allowing you to adopt best practices.
If you simply want to automate your Excel, there are other alternatives. But realize what you are missing out on:
These are two different decisions
In both cases, that experience and day-to-day knowledge are extremely relevant, and they must be communicated so that the business rules are configured correctly.
Furthermore, even though we may have many clients in the same industry, each client has their own particularities: they view information differently or manage it at a different level. All that feedback is incredibly valuable because it allows both teams to align and achieve results faster.
"Great, I have artificial intelligence and deep learning in my supply chain now, so I want to get long horizons and things I have never projected manually"
It is common for people to arrive with the expectation: "now that I have a specialized platform, I want to forecast everything I can, with infinite granularity." And in practice, or at least statistically speaking, it doesn't work that way.
So, they want to go from one or three months to projecting 12 or 24 months. And that is when you start to ask: Come on, do you really need it? How are you going to make the most of it? How are you going to work with it if you aren't used to doing it that way?
Or I’m going to switch from forecasting at the SKU level to the customer level, ignoring the fact that while the more disaggregation there is, the greater the deviation and uncertainty in the forecast accuracy.
The same applies to longer horizons. So, if I decide to stretch the horizon while simultaneously taking it to the most granular level, I’m introducing a ton of uncertainty right from the start, and in a process I’m not used to handling. Naturally, when they start receiving the results, they don't know what to make of them, how to interpret them, or how to integrate them into the process.
My recommendation is:
Start exactly how you’re doing it today. It’s the best way to begin because you’ll build a learning curve much faster. Once you have that stabilized and see that you’ve improved compared to where you were, then it’s worth taking it to the next level.
It’s either I dive in and explore, potentially crashing because I didn't listen to the recommendations of a company that has been doing this consistently across different industries, or I take advantage of those recommendations and experiences. Because if this is my first project and I’m moving from Excel to a platform, it’s like trying to fly before I’ve even taken my first steps. These are decisions, and it’s about how you want to approach your learning process.
"We’ve already implemented the platform, but I need to audit the data it generates. What’s the best way to do that?"
Sometimes double work happens because planners don't fully trust the platform's data and can't let go of control: they see the platform but keep calculating things manually. The question is: As a project leader or supply chain manager, how do I convince my analysts to stop doing manual calculations?
You don't just switch from one to the other andcall it a day. You always see both systems coexist for a period. The question is how long you want to drag that out and how practical it is, because I could spend all my time on validations and perhaps keep finding things.
There are different tools to shorten that period.
Control items. This is one of the first things we work with to ensure we are looking at the same data. They allow you to narrow down the sample and identify specific products to monitor and compare.
The ranking. Within the platform, we provide a ranking that helps you prioritize: knowing what to pay more attention to and where to start validating. That’s where you’ll find products with similar patterns, such as high-revenue items, which you always want to keep a close eye on.
Start with high-frequency, stable-behavior items, which are very easy to validate because they are the most consistent, and leave the variable-behavior ones for later. Once you understand that dynamic and shift your mindset regarding stable versus variable products, or high-impact versus low-impact ones, your approach to analyzing information changes as well.
The validation sequence
Always look at the general values and compare three things: what actually happened, what was projected, and what I see in my system. From there, drill down:
If I see that those areas are fine, it’s time to start migrating. Because trying to keep doing it both ways definitely doesn't make sense.
"I'm not going to do anything anymore" or "my work is no longer valuable".
Some planners or analysts feel threatened by the arrival of these platforms: "I'm not going to do anything anymore" or "my work is worthless now". How do you make it understood that your judgment isn't being replaced, but is actually more important than ever?
You see it in cases as simple as generative artificial intelligence. You don't just settle for "this is what it told me and that's it". Judgment will always be relevant because it allows for those validations; it allows you to know that things are correct.
And once they are correct, with judgment and experience comes the important part:
So that it isn't just another report within the company, but rather that those decisions start generating results: higher sales, savings, lower costs. Let's focus on this branch or this product group and start iterating more.
That's where you have data-backed support to validate hypotheses, execute them, and see the results. And it's iterative: the more you get involved and the more you understand it, the more possibilities you will find.
That happened to me a lot even when I worked with Excel, because it was always about analyzing and reviewing data, and based on that, seeing what is working and what isn't. It's just that you spent all that time on operational tasks.
When you start to question things a little and look at data in a different way, that's when you start to find new strategies and see opportunities for optimization. That is where the real value lies, and with your experience as a planner and data expert, that is where you can contribute.
Instead of being the owner of the formula ("this is how I do it"), the real value lies in the conversations generated within the team through the numbers. Because numbers become triggers, and you can quickly validate whether a hypothesis had an impact on demand: did we move more products, did we align better, did we negotiate better terms with suppliers?
Instead of putting out fires, you can focus more on results than on the day-to-day. You become more proactive in your execution. And that is where happy problems and growth will come from.
Supply chain is one of the industries where artificial intelligence has yet to fully deliver: according to Gartner, 55% of supply chain directors are unclear about the return they are getting, even though they already allocate 67% of their digital budget to it. And I believe this is precisely why: we have the data, but not the full context.
And because it is such a complex operation (demand planning, inventory, purchasing, promotions, seasonality, and specific sectors that you as a planner already know well), you may have that information while the tool does not.
The tool is a support system, but the decision is yours. If an external event occurs, you already know it will affect your operation; the tool didn't get the memo, it doesn't say "this news happened, I'm going to reduce purchases." That comes from you.
When you think about the threat of a platform replacing you: no, that moment hasn't arrived yet and it won't arrive anytime soon. What you must do is leverage these tools so you don't get left in the past. Just as Excel arrived to replace manual files and long, extensive spreadsheets, saving us a lot of time, these new tools are now here to save us even more time.
In all our modules, both in planning and inventory, we allow the client to make adjustments. Why? Because it is human knowledge that adds value and feeds back into the entire process.
When someone tells me "I want to automate my process 100%", that's when you tell them:
Regarding demand planning. There will be variability that the tool won't be able to capture (negotiations, marketing campaigns), and it's important that you provide feedback within the platform. Generate those inputs so it can keep learning and provide recommendations based on them.
Regarding procurement. You can have your reorder point, you can adjust it, and you will always have the recommendation, but there are also products that deserve a little more attention. That's why we always build it based on ranking: it allows you to know what to prioritize and better understand the diversity you have with a supplier. In the order you are about to place, what needs to be guaranteed? Because that will impact service levels, the income statement, and the company's profitability.
It's a huge help, because it means having that information already processed. But that knowledge and those decisions are what ultimately close the loop and add all the value in the world.
Are you in the middle of this transition or thinking about making it? Write to us and let's talk about what it looks like for your operation.
Laura Ramírez is a business and operational planning supply chain strategist with more than 15 years of experience helping companies make better decisions based on data.
.jpg)
