How To Modernize Your Data Strategy And Infrastructure For 2025

We are still in the early days of data and the value it can add to companies.

You’ll read plenty of statistics about how much value data can drive and how far behind companies that aren’t using data are. And as a data consultant, I have helped companies find that value in their data.

It does exist.

Regardless of where you are in your data maturity journey, it’s important to realize that it takes time to build up data infrastructure as well as a data strategy and processes that can take your company to the next step.

You need buy-in from leadership to invest in new data strategies and initiatives, and that’s far from easy, especially going into 2025, which seems to be an uncertain year.

So as the new year approaches, it’s important to take a moment to start planning out how you can take your data infrastructure and strategy to the next level. Just because the economy is uncertain, it doesn’t mean your data strategy and projects can’t drive value in uncertain times.

For this article, we’ll be discussing how your data team can find those impactful data projects that are aligned with the business as well as outline several examples of projects we have seen valuable that your data team can implement.

So let’s dive in.

Align With The Business

Data teams need to be aligned with the business, and what the business strategies are overall.

I’ll keep this section brief as I have several articles discussing the importance of this subject.

The too long didn’t read part of this section is that your data team’s projects need to either clearly align with what the business is doing and their strategy, or the project must have so much value that the business would have to stop what they are doing and realign.

In order to align yourself with your business stakeholders, you need to talk to people. I know, it’s not everyone’s favorite thing to do in the tech world.

But when I talked with Ethan Aaron from Portable, he discussed how there wasn’t a central data team when the company he had worked for got acquired. So suddenly, his new role was Head of BI. The first thing he found out as he went out and talked to teams to figure out what needed to be done was that marketing already had its own data tools, as did sales and operations. Of course, the departments rarely shared data with each other, missing out on massive opportunities.

So one of the first major goals he realized needed to occur was centralizing the data to allow users to have access to multiple data sets, reduce duplicate costs and work when it came to dashboards and reporting, and create a better alignment with data and the business.

All of which started with talking.

On a similar note, when I worked at Facebook, we’d spend time at the end of every half talking with our stakeholders to understand a few points:

• What problems they were dealing with
• What projects they were taking on
• What projects they think your team should take on?

But we didn’t just ask questions; we also would put together a deck to cover some of our past projects as well as what we had already thought about in terms of projects we could take on.

This provided a tool for everyone to review and be inspired by. You shouldn’t just go asking for work, but also, sometimes, your stakeholders might need some help crystalizing the work they should be taking on.

In the end, this whole process should get you a clear set of pain points and projects that teams want solved.

Find Projects Worth Doing Not Just AI

Let’s make this clear–you need to find data projects worth doing. In fact, maybe you shouldn’t even look at them as data projects per se but just overall business projects. Projects that push the needle, and in the world where everyone is talking about AI, it can be tempting to lean on only finding AI projects.

After all, your boss is probably pushing you to do so. Because then they can tell their boss or board of investors that we are using AI. I literally had that conversation with another data leader just a few weeks ago. Where they were being pushed to have some AI story, for no other reason than to have an AI Story.

So let me give you some advice. As someone who has helped dozens of companies set up their data infrastructure and drive value via data, don’t get too distracted by AI.

Don’t get me wrong, we’ll be talking about AI later in this article, and there is value in using it. But you need to find projects the business can use. Think about it, if you build an AI model that perhaps puts out interesting results, maybe even about your customers, but you have no way of implementing it, what is the point?

Examples of Data Projects You Should Take On In 2025 

There are dozens of projects you can take on as a data team. You can migrate data stacks, you can try to help reduce churn, etc. So here are a few example projects I have led and know can drive the needle.

Projects that don’t just slightly reduce costs or improve revenue. Instead, you’re looking for big wins.

Reduce Costs

Reducing data infrastructure costs can be a great project. I want to emphasize “can” because the truth is that many companies want the data teams to be a value add. And yes, obviously you can save tens if not hundreds…ok and in some cases millions of dollars from your data infrastructure.

But it can also feel as if you didn’t do much (or perhaps you saved money that should have never been spent).

That preface aside, cost savings projects are great.

They can be used to get a budget for other projects that are going to increase a company’s bottom line. So I do enjoy starting with cost savings projects as they can be a good stepping stone to further projects. Once you’ve saved $50,000, now you can ask for $25,000 to try out something that could return $500,000.

Personally, I find that looking for some initial big wins early on in the year can help set your data team up for further success. One of those can be to cut costs.

In the end, it’s not the real win you’re trying to go for, though.

Customer Focused Projects

Every industry has a different customer, with different concerns, behaviors, and issues. But every industry must care about their customers.

After all, customers are the life blood of a business. If they aren’t happy or maybe if there is a chance But there do tend to be a few projects that I seem to see a lot.

Reducing Customer Churn 

Whether a service is digital or not, there is a lot of benefit in knowing if you lost a customer. Whether you’re calling that churn or a more specific term for your industry, it’s worth figuring out how to improve this KPI.

In order to start improving this issue, you’ll need to ask questions like: Are you losing customers in certain parts of a region? Are they of a certain age group? AndAre people only happy for 60-days with your service then quitting?

If you can spot these patterns, you can work to improve your issue and, in turn, increase your top line.

Detecting Abnormal Customer Behavior 

This could be fraudulent behavior, super users, confused customers, etc. For many clients I have worked with, there is often a need to detect users who have abnormal behavior. This can help target either behaviors you don’t want to see, or perhaps behaviors you’d like to encourage. For example, I have now worked on several fraud projects that have saved customers millions of dollars. In many cases, some companies will miss bad behaviors and, in turn, their customers will continue doing them.

Customer Segmentation

A simple project many companies can take on is a customer segmentation project. That is to say, you may have customers that are driving 80% of your business, and you might not know it. You could be focusing most of your marketing dollars on said segment but instead, you’re burning it on too broad of a market. Similarly, you might have a small market segment that is starting to grow, and if you put in a little more focus either on the product or marketing, they might become a brand new large segment. But this all starts with you understanding your customer.

AI Projects and How To Handle Them In 2025

“Most Companies Want To Do AI. Most Are Barely Doing BI.” – Joe Reis

Look, it can be hard to want to do a basic BI project when it seems like everyone else is doing AI. And I do like that Joe Reis pointed out in his article that AI and BI are not the same skill sets. You can take on AI projects without having BI.

However, I do believe that if you don’t have a solid foundation built in terms, knowing where your data is, how you’re going to store it and manage it, it’ll be really hard to take on an AI project.

But if you’ve got a reliable data infrastructure and have proven that you can deliver basic data projects, then I do believe it can be worth taking on an AI/ML project…assuming there is a valid business case.

Finding Business Processes That Aren’t Working

I once was talking to Gordon Wong, who is currently the Head of Data & AI at Newfire Global, and is an ex-data VP of both Fitbit and Hubspot, and he talked about some of the projects that have driven value in his experience. One that isn’t as exciting but can be very impactful is knowing what to stop.

This could be that ad campaign that you’re spending millions of dollars on, or it could be a business arm that really isn’t profitable.

In one example provided by Gordon, he discussed how he worked at one company in which the metric that the marketing team used to judge their performance was the number of letters/emails sent out.

I assume most readers can see the issue here. This metric is based on input and not some form of business output. This also cost the company hundreds of thousands of dollars because they were sending letters to people on their “do not send list” (along with other implications).

As a quick call out, one point Gordon made was that one of the lessons he learns over and over again is the importance of:

Finding the real metric – Gordon Wong

In this case, by shifting the discussion to what the real metric should be, he saved the company several hundred thousand dollars just by turning off things that weren’t working.

You do always need to be careful here as you will be stepping on the toes of other directors and leaders. So make sure you don’t blindside any one. I’d recommend you bring them in and help them also get the win (of course, make sure you get credit).

But work with them to bring the ideas to your CFO, CMO, or whoever is going to make the final call so they see it was a joint effort.

All of this being said, let’s talk about the flip side; instead of talking about the data projects you should take. Let’s talk about the data projects you need to avoid.

What Your Data Team Should Avoid

Now that we’ve talked about the projects you should take on as you’re leading a data team,  talk about a few things you should avoid as they can risk you not delivering value with data or possibly losing your job.

Lack Of Ownership 

One theme that tends to come up a lot, even at large tech companies, is a lack of ownership of core data sets and pipelines.

Whether it be Airbnb or several clients that I have worked with directly, not clearly identifying who should be responsible for pipelines tends to cause a lot of distrust as well as the rebuilding of the same data sets over and over again. I have seen this happen with several clients. In fact, at one project I came in, no one was even using the dashboard because so much distrust had come over the pipelines and the entire data warehouse.

Why?

There were no owners.

They were just tables without a home.

All of which can be fixed by either creating a process that ensures teams don’t set-up tables without ownership or you set up technology to capture ownership.

Infrastructure For Infrastructure Sake

For every company spending two years trying to set up the perfect data stack, there is one using a Postgres replica to answer their business critical questions today. Don’t let all the architecture diagrams confuse you; most companies I have either worked with or talked to have built things in an iterative way. Meaning, they didn’t go out and spend hundreds of thousands of dollars on the most expensive data solutions out there.

They proved that they could deliver reliable data projects. And slowly they built out data infrastructure that met their needs. In fact, in some cases, I’d say Facebook uses less tools than some SMBs who feel like they have to sign up for a data catalog, a data quality tool, a data asset management solution, etc.

ML First, Analytical Processes Never

This is somewhat attached to what we’ve discussed already. Many data teams will run towards AI and ML without ever taking on any BI projects. In fact,  I have now had 3-4 discussions with companies that are running towards machine learning without any form of core data infrastructure or any plans to build it out. What this generally leads to is a lot of one-off data pipelines built in Python notebooks that are more manual than they are automated. This leads to unmaintainable systems.

I know everyone seems to be ahead of you on the data side. But if you don’t have some practice working with data, you’ll have a hard time trying to take on AI and ML projects.

Find The Right Tools

To end on a final note, that perhaps makes more technical individuals happy, finding the right data tools can make your life easier. After all, if you talk to most data experts, they will all say you need to consider people, technology and processes when trying to set-up a successful data strategy.

So although tools can be distracting, picking the right data solutions means your team can spend less time rewriting the same data connector or redesigning SSIS or Airflow.

So do take a moment to pause and ask yourself, are their data solutions that could help us avoid spending the next month or more to simply move data from point A to B and does it fit our teams skills and current infrastructure.

If you need help picking the right data solutions, feel free to set-up a consultation here.

Planning Your Data Strategy 2025

Data will continue to provide value well into the future. It can help your company reduce costs, both internally on your own data team as well as finding cost savings opportunities in general. It can also help your business drive new value by improving how your customers experience your service.

However, all of this will require a mature data infrastructure and set of processes. This does mean you might occasionally need to push back on management to ensure you build reliable data infrastructure.

If you’re still hoping to learn more about this change and some of these skills, you can check out these articles.

Common Pitfalls of Data Analytics Projects

9 Habits Of Effective Data Managers – Running A Data Team

The Data Engineer’s Guide to ETL Alternatives

Build A Data Stack That Lasts – How To Ensure Your Data Infrastructure Is Maintainable

Explaining Data Lakes, Data Lake Houses, Table Formats and Catalogs

How to cut exact scoring moments from Euro 2024 videos with SQL