What to Ask an Intent Data Provider Before You Buy

Intent Targeting · Published 2026-08-20

Ask five things before you sign with an intent data provider: how they build the topic taxonomy, how often the data refreshes, how they de-anonymize visitors, what regions they actually cover, and how they calculate a surge. Get vague answers on any of those and the data will look busy in your dashboard and do nothing for your pipeline.

Start with the intent data providers question nobody asks first: what is the topic taxonomy?

Every provider sells “topics,” but the taxonomy behind them is where the differences that matter actually live. A narrow, keyword-scraped taxonomy tags a company as showing intent on “marketing software” when they were reading a listicle that mentioned it in passing. A properly built taxonomy maps topics to a category tree with defined parent and child relationships, so a surge on a specific sub-topic (say, “content syndication vendors”) means something more precise than a surge on “B2B marketing.”

Ask the provider for the taxonomy document itself, not a sales deck screenshot. Ask how often it is revised and who reviews additions. A taxonomy that has not changed in two years in a market that changes every quarter is a sign the vendor stopped investing in the data layer after they closed you as a customer.

Ask how the refresh rate is measured, not just stated

“Weekly” and “daily” both show up on data sheets, but the number that matters is how long it takes a real signal, a prospect actually researching your category, to show up as a scored account in your CRM. That includes the crawl or panel collection window, the processing and scoring pipeline, and the sync to your stack. A provider that refreshes its raw data daily but only pushes scored accounts to your CRM once a week has a weekly product no matter what the homepage says.

Ask for the end-to-end number in writing: signal observed to account visible in your tools. If they cannot give you that number, they have not measured it, which tells you something on its own.

Why this matters more than it sounds once you are comparing vendors

By the time a buying committee is actively comparing vendors, the window where outreach still looks helpful instead of late is short. A provider whose real latency is ten days is handing you accounts your competitors already called.

Get specific on de-anonymization method

“De-anonymization” covers a few genuinely different techniques and providers rarely volunteer which one they use unless asked directly.

IP-to-company mapping: reliable for office traffic on a stable IP range, weak for remote and mobile, and increasingly weak as more traffic routes through VPNs and shared cloud IPs.
Cooperative or panel data: the provider is part of a network of publishers or a data co-op that pools observed behavior across member sites, which trades some precision for much broader coverage of what a specific account is reading, not just visiting.
Device graph or identity resolution: probabilistic matching across devices and browsers using third-party signal, which is the most granular when it works and the most exposed to the ongoing decline of third-party cookies and cross-site tracking.

None of these is universally correct. What matters is that the provider tells you which one (or which mix) underlies the score you are buying, and what their match rate actually is on your specific target list, not on their best-case reference customer.

Confirm regional coverage against your actual territory, not their homepage map

A provider’s coverage map on their marketing site almost always shows the whole world in the same shade. Ask instead for the count of active publisher or panel sources in the specific regions you sell into. A vendor with deep US coverage and thin coverage in EMEA or APAC will still show you scored accounts in those regions, they will just be scored on far less signal, and you will not be able to tell the difference from the dashboard alone.

If your pipeline is split across regions, ask for a sample export filtered to each one before you buy, not after. This matters most for teams expanding into a new territory: the provider that looks strong in your existing pipeline dashboards can quietly be running on a handful of publisher relationships the moment you point it at a market you have not sold into yet, and you will not find that out from a contract renewal call six months in.

Understand how surge is calculated before you trust it

A “surge” alert means an account’s research activity on a topic jumped against its own baseline, but the baseline window and the sensitivity threshold are provider decisions, not a fixed standard. A provider using a short baseline window will flag more surges, including noise from a single research session. A longer baseline is more stable but slower to catch a genuine buying-cycle start.

Ask what the baseline period is, whether it is set per account or per topic, and whether you can adjust the sensitivity. A surge feed you cannot tune to your own sales cycle length will either flood your reps with false positives or miss the window entirely. A short sales cycle wants a tighter, faster-triggering baseline; a longer enterprise cycle wants a wider one, or the same account will trip the alert every few weeks on ordinary research activity and reps will start ignoring the feed altogether.

What good looks like when you put the answers together

None of these five questions has a single correct answer, because the right taxonomy depth, refresh cadence, de-anonymization method, regional weighting, and surge sensitivity depend on your market and your sales motion. What a credible provider has in common is that they answer all five with specifics, in writing, before you sign, instead of pointing back to the homepage. If a vendor cannot produce their own taxonomy document, latency number, de-anonymization method, source count by region, and surge methodology on request, treat that as the answer.

We run intent programmes for clients across all four of these variables, which is exactly why this list is the one we wish more buyers asked before signing. See how we evaluate and activate intent data and buyer signals for a full-funnel programme, or start with the difference between first-party and third-party intent data if you have not settled that question yet. If you are new to the category, what B2B buyer intent data actually is is a good place to start.

Browse the B2BinDemand library for more intent data guides and buyer checklists like this one