How we used Demandbase AI to build our new EMEA Sales territories

There are spreadsheets, account lists, Closed Lost reports, Closed Won reports, industry analysis, and inevitably someone asking why this account is with that rep.

At some point, you usually end up with a spreadsheet called something like EMEA_TERRITORIES_FINAL_v7_ACTUAL_FINAL.xlsx.

As an SDR Manager, one of my biggest priorities is making sure the team is focused on the right accounts, at the right time, with the best possible chance of turning activity into qualified pipeline.

So when we started building new territories for our New Business EMEA sales team, our Sales Leader and I wanted to approach it differently.

Instead of starting with a massive account list and manually dividing it between reps, we used Demandbase AI to analyze our market, challenge some of our assumptions, and help us build our target account lists.

We didn’t ask Demandbase AI to build the territories for us. We used it to work through the decisions with us, one prompt at a time.

Here’s what that process looked like.

Step 1: Start with what you already know

Before deciding where we wanted to go, we wanted to understand what had happened previously.

We started with two simple prompts:

“Can you show me the Closed Won & Lost data across industries?”

Followed by:

“Can you also show the Closed Lost reasons?”

That gave us a useful baseline. Where had we historically created opportunities? Which industries appeared most often? And why were we losing deals?

The reason behind the loss mattered, too. An account that went Closed Lost because the timing wasn’t right is very different from one that simply wasn’t a good fit, so we didn’t want to treat them the same way.

Step 2: Move from reporting to strategy

Once we understood the historical picture, the next question was what we actually wanted to do with it.

Our next prompt was:

“Based on all of this, how should we shape our target account lists in terms of industry spread?”

Then we looked specifically at accounts we’d previously worked:

“What Closed Lost accounts would you include? Only select ones without journey stage = SQL, pipeline or customer.”

And tightened the timeframe:

“Can you include Closed Lost from the last 18 months?”

That turned Closed Lost into another useful territory-planning signal.

We could start identifying accounts that genuinely deserved another look, rather than filling territories with every company we’d ever spoken to.

Step 3: Give AI your strategy

One thing became clear pretty quickly: AI is much more useful when you give it your strategy instead of asking it to come up with one for you.

Our Sales Leader and I already had views on where we wanted to increase or decrease coverage, so we put those directly into the prompt:

“Using the industry recommendations, generate a recommended EMEA Target Account List for the upcoming quarter. The objective is to maximise pipeline creation and revenue by prioritising accounts with the highest likelihood of becoming opportunities over the next 6–12 months.”

We also added:

“The recommended account mix should reflect these priorities rather than historical allocations.”

We wanted historical data to inform our decisions without allowing it to dictate them.

Otherwise, it’s easy to use AI to create a more sophisticated version of the territory model you already have.

Step 4: Define what makes an account worth a rep’s time

Industry alone obviously isn’t enough.

Two companies in the same industry can have completely different likelihoods of becoming pipeline, so we gave Demandbase AI our account-ranking framework.

The prompt started with:

“Within the priority industries, rank accounts using the following factors…”

We broke that framework into three areas.

1. Buying readiness — highest weighting

We asked Demandbase AI to prioritize accounts that were Pre-SQL and showing signals such as high engagement, high intent, increasing intent over the previous 30–90 days, recent website activity, and multiple engaged buying group members.

For us, it came down to a fairly simple question:

Is something happening inside this account right now?

2. Fit

Next, we asked it to prioritize accounts matching our ICP and firmographic requirements: based in EMEA, with the right company size and revenue profile, and aligned to our ideal technographic profile.

Intent without fit only gets you so far. And a perfect-fit account showing no buying activity isn’t necessarily where a rep should spend their time either.

We wanted to see both.

3. Pipeline history

Finally, we asked it to prioritize:

Accounts that had never been worked but were showing strong engagement

Recycled opportunities showing renewed intent

Closed Lost opportunities from the previous 18 months where the reason was Timing, Budget Delayed or No Decision

That gave us a better way to think about recycled accounts.

Instead of writing an account off because we’d already tried it, we could ask:

“What’s changed since we last tried?”

Step 5: Clean the list

Anyone who has built a target account list knows generating accounts is only half the battle.

You then have to remove everything that shouldn’t be there.

We started simply:

“Can you remove accounts that are customers?”

Then:

“Can you also remove accounts with ultimate parent HQ = US, Canada or South/Central America?”

The ultimate parent rule was especially important for our EMEA territory model. We didn’t want a local subsidiary appearing in a territory if ownership of the ultimate parent meant the account belonged elsewhere.

This became the working pattern throughout the exercise:

Generate → inspect → exclude → refine → inspect again

We weren’t trying to come up with one enormous, perfect prompt. We were working through the data and refining the answer as we went.

Step 6: Now build the territories

We didn’t start assigning accounts to reps until we had the account pool where we wanted it.

By that point, we weren’t dealing with a giant list of EMEA companies. We had a prioritized account pool informed by:

Strategy + historical pipeline + ICP fit + Demandbase Intent + engagement + journey stage + account hierarchy

Then we could give Demandbase AI our actual territory rules. The specific rules aren’t really the important part, what mattered was the order in which we applied them.

We had already done the strategic analysis before we started assigning ownership. We were dividing up a curated market, not just dividing up a database.

Step 7: Pressure-test the account mix

Once the territories were built, we wanted to make sure they still reflected the industry strategy we’d agreed.

We gave Demandbase AI our recommended H2 2026 industry split and used that as a final sanity check.

Were we over-indexed in an industry we had deliberately decided to reduce?

Had one territory become disproportionately weighted toward a particular sector?

The percentages weren’t there to override strong account-level signals. They gave us another way to pressure-test the final territories and make sure the overall mix still matched the strategy.

Step 8: Make it usable

Eventually, you have to stop analyzing and give your reps some accounts.

Our final prompts were probably the simplest of the entire exercise:

“Can you share the full list of accounts?”

And:

“Can you create an exportable list of each?”

And that was it.

We’d moved from historical Closed Lost data to a prioritized, strategically weighted, rep-level EMEA target account list.

What I took away from the process

The biggest benefit wasn’t that AI magically created our territories. It didn’t.

What it gave us was a better way for my Sales Leader and me to work through the territory together. Instead of defending our own lists, we could interrogate the same data and ask the same questions:

Why is this industry overrepresented?

Which Closed Lost accounts deserve another look?

Which accounts are showing buying signals now?

Will these territories give our SDRs and AEs the opportunity to reach quota?

What happens when we apply our customer and parent-company exclusions?

Does the final account mix actually match the strategy we agreed?

And ultimately, why does this account deserve this rep’s time?

That was probably the most valuable part for me as an SDR Manager. Instead of creating an “SDR list” and handing it over to Sales, we were building the territory together.

Don’t start with “build me a target account list”

If I were doing this again, that’s the one piece of advice I’d give.

I wouldn’t start with:

“Give me the best 900 accounts in EMEA.”

Start with your historical performance. Understand why you’ve won and lost. Decide where you strategically want to invest. Define what buying readiness looks like. Add your fit criteria. Apply your exclusions. Then introduce your territory rules and pressure-test the result.

Each prompt makes the next prompt smarter.

By the end, you have more than a spreadsheet of accounts. You can explain why the territory looks the way it does: where you’ve historically seen opportunity, where you’re deliberately investing, what Demandbase is telling you about buying readiness, what you’ve excluded, and why those accounts are worth a rep’s time.

That’s a much better starting point for a new territory, and considerably better than EMEA_TAL_FINAL_v7_ACTUAL_FINAL.xlsx.

Build smarter territories with Demandbase

Book a meeting to see how Demandbase helps sales and marketing teams prioritize the right accounts, act on buying signals, and focus reps on the opportunities most likely to turn into pipeline.