How To Preserve Your Data Quality (And Your Sanity) Through A Merger, According To A Revenue Leader Who Aced One

Amy Palmer’s path to marketing leadership was an unconventional one. After four years investigating police misconduct in NYC, she did a stint as a manuscript editor and earned a master’s degree in anthropology before transitioning into demand generation marketing.

Today as Head of Revenue Marketing at Autodesk Construction Solutions, Amy attributes her success to her ability to seamlessly inhabit different worlds. “One thing that’s always been really important to me,” she says, “is embracing multiple perspectives.”

When her previous company PlanGrid was acquired by Autodesk in 2018, it was joined by two other recent Autodesk acquisitions BuildingConnected and Assemble Systems. Amy was suddenly faced with three tech stacks and data structures that she had to blend with a fourth platform — Autodesk’s existing BIM 360 ecosystem—with the daunting end goal of creating a single marketing tech stack to support a brand new business unit.

While the gargantuan task of merging multiple Marketo and HubSpot instances is enough to make even the most seasoned marketing operations professional recoil, Amy’s diverse skillset was key to pulling it off. With an editor’s meticulous attention to detail, an anthropologist’s understanding of how people use systems, and an investigator’s intuition, Amy successfully led her team to re-architect the company’s entire data ecosystem.

In this interview, Amy reflects on her experience undergoing a merger, and shares her triumphs, battle scars, and a few particularly hard-won lessons.

Nick: What’s the #1 lesson you learned from going through a merger?

Amy: You can’t know the future, but you can save your future self a ton of grief by keeping your data as clean, usable, and transformable as possible. Back at PlanGrid, we had a clean Marketo instance that we started from scratch. We were very clear on the way we wanted to look at the world and our go-to-market model. 

Three years later, everything has changed. Suddenly we’re part of Autodesk and selling in ways we never would have thought possible. There have been so many times we’ve been glad we maintained certain data structures, because changing external circumstances suddenly made them business critical. That’s why future-proofing your data in a granular state is one of the smartest things you can do, even if you don’t necessarily see a use case for it now. You can always aggregate it later if needed.

What are the most common causes of bad data in your experience?

Bad data can creep into the system in so many ways, so you need to keep watch on many fronts. 

Field duplication issues are common—this was especially challenging when we had to merge multiple Salesforce instances on a time crunch. Poor data hygiene is another common one, often resulting from prioritizing shiny new development projects over far less glamorous data maintenance projects. Bad data can also emerge from not following processes, like when business development teams make calls but don’t log them, or when marketers misuse program templates. 

But bad data can often be more a symptom than a problem. Investigating the root cause of bad data—whether inefficient processes, internal misalignment on goals, or tech stack shortcomings—is often the key to eradicating it, and finding business process improvements in the process. Too often we fix the impact of problems without slowing down enough to really understand why they are happening. 

How do you balance speed with accuracy when you’re merging three Marketo instances?

Both are essential, yet trade-offs are inevitable. Dividing the transition into phases is critically important, as well as setting expectations with your team. I had to prioritize the most critical business processes to support first, and create a plan to tackle the longer-term processes over the course of several months. It can be hard to stick to the plan, especially when teams are adjusting without the tools they’re used to having. But you have to remember you prioritized for a reason: So you don’t lose focus of what will help the business the most.

Since data is used by multiple groups within my organization, our process involved input from a wide variety of stakeholders. If one team thinks we’re going to be able to provide a whole set of reports in month two, and we’ve scoped it for month four, aligning on those expectations early on is essential.

If you could go back in time, what would you tell pre-merger Amy?

We thought a lot about the process of merging our data, since it’s such a huge undertaking, and a little less on what our ideal data end state was. In some ways it’s inevitable, because there’s just so much work to get it done—we were migrating Marketo programs, nurture campaigns, customer data. We were also redefining people’s roles, training new teams, and trying to avoid slowing down our growth engine during the whole thing. It’s just immense. But to be candid, it’s taken longer than we expected to build back to the level of insight that teams had prior to our merger, and in retrospect I think we should have built that more into the primary scope of the project.

The project was to integrate first so we could stay executionally stable during the transition, and then figure out how to report after. In retrospect, the project should have been: How do we blend processes to drive the outcomes we need? It seems subtle, but I think mentally framing it this way from the start would have been a big time saver.

What’s one thing you’re doing differently with your data this year versus last year?

Our company as a whole is aligning around data better, and we formed a new data team this year. One of the great things about a bigger blended team is that we now have resources to help align on a view of the world, or rather, multiple views of the world tailored for various teams — for example, executive-level metrics, board-level metrics, and CEO-level metrics.

When I used to be solely focused on marketing ops, for example, I never had access to data science resources, so my perception of what’s possible with data came from what I could access through Marketo and Salesforce. And there are some pretty sophisticated concepts behind the reports in those platforms. But when you start to think about custom-baking predictive algorithms for your next white space in your account, or which customer is likely to buy a product based on past product usage, your world really opens up.

How do get technically-minded data analysts and business-oriented marketers on the same page?

It’s a great question since they really are such different but equally critical mindsets. To stay aligned, we’ve set up a cross-functional pod we’re calling our “marketing metrics tiger team,” where representatives from marketing, inside sales, sales ops, and our data team are focused on prioritizing and executing all the reporting requests coming in from the extended team.

I felt it was necessary to have representation from the data team sit in on all strategic meetings so they can glean insights on how the marketing team uses metrics to understand the business. And business-oriented marketers are learning a lot about how data needs to be structured and transformed to enable business reporting. We’re finding a ton of value in helping the teams find a common language.

Any unexpected challenges?

We underestimated how hard it was to merge different systems, even when they were fundamentally looking at the same type of data. One of the groups that we brought in had a HubSpot instance and their own BI dashboard. They had access to great attribution data, but how they consumed and used it was vastly different from how we did, so merging our processes was more difficult than we thought.

Enablement has also been a much bigger lift for this whole process than we ever thought it would be. It’s easy to get caught up in thinking: How am I going to map these leads source fields to these leads source fields? And that’s certainly a challenge, but actually the hardest part is developing business alignment around processes with key stakeholder groups.

What’s your top tip for improving data quality?

Ensure your stakeholders’ priorities are aligned. It sounds obvious, and it’s great to embrace  in theory, but it often falls to the wayside, especially when you have many smart, action-oriented people working to solve multiple urgent projects in parallel. If you aren’t in agreement about the problem and why solving it is important, and if you don’t understand your stakeholder’s priorities, work tends to get endlessly tweaked and redone, rather than actually finished and used in a way that can impact the business.

About the authorNick is a CEO, founder, and author with over 25 years of experience in tech who writes about data ecosystems, SaaS, and product development. He spent nearly seven years as EVP of Product at Marketo and is now CEO and Founder of Syncari.

Join hundreds of revenue leaders fighting the good fight against bad data.