Tl;dr - Whether you call it purchase intent data, buyer intent data or just plain old intent data, there's a common misperception that the preponderance of the data signals indicates bottom-of-the-funnel research. That's a problematic assumption most common among sales teams. In aggregate, signals can be really helpful, but reading too much into any specific signal is risky. That's an opportunity for sales operations and/or marketing to address through sales enablement.
Will purchase intent indicate who's researching?
The short answer is not necessarily!
Despite all the contrary claims that float around out there.
Longer answer - it will help you understand.
Here's why.
Is Every Visitor to Your Website Researching?
Let's start with your own website.
You probably have a visitor-to-lead conversion rate of 2-5%. Some people who visit and don't convert are researching. But others wander along because your SEO and inbound marketing content efforts answer questions people are asking.
Some people just come for those answers. They're not researching. They're definitely not buying.
If that's true on your site, it's true elsewhere. Especially when we don't even really know whether they're looking for an answer!
Understand the methodology
So then it's important to understand how data providers collect and report signals.
Bidstream Data
Bidstream data is "intent data" which is a byproduct of programmatic advertising platforms (DSPs or demand side platforms) and which is derived from observations of online ad interactions. It's based on a large leap of faith - that someone who is shown an ad, is seeing it adjacent to content the advertiser hoped was relevant, and therefore that the person shown that ad is researching the topic of that content. Let's put aside the issues we know exist with matching content and ads, and let's further assume that the content is pertinent and that the person actually read it.
That means:
- If someone was shown a certain ad, it's assumed that the content on the page they visited was somehow related to that ad
- If they were consuming that content, it's assumed they were specifically interested in that topic
- If they were interested in that topic, it's assumed they were researching it
That's a stretch.
What you know is that someone was shown an ad. From that ad impression, you can possibly surmise that the content was related.
That doesn't mean the natural language processing (NLP) which determined the content topic is accurate; it doesn't mean they are where they want to be; and it certainly doesn't directly equal "research" or purchase intent.
Publishers' Coop Data Exchange
While Company Surge® sounds like "search" when said quickly, this data actually indicates above-average consumption of content, that's related through an opaque taxonomy to topics, by individuals in a company/location. (Only Google, Bing, etc. know who's conducting what internet searches.)
Because publishers own the content they publish, they are able to tag content directly during the editorial and publishing process rather than rely on NLP. Therefore it's likely a reasonable association to the topics they tag. Further, because it's the publishers' first-party data, they're able to observe engagements including site search, page views, and scroll percentage.
This is a more reliable way to understand engagement (time, scroll, number of pages, etc.) across the range of cooperative publisher sites.
But does engagement equal research?
No.
And does it capture all the critically important signal beyond the small range of coop sites? No.
An intent data syllogism might posit that all (OK a lot of) purchase intent will appear in good intent data, but not all intent data reflects research, much less true purchase intent.
Research Doesn't Mean You're (They're) Buying
Like you, I research target companies during the sales process, while considering investments and in the course of considering independent director board roles. I also research companies and products when I'm interested in them. And of course, I research when I'm buying.
The point is that intent data is an important signal. It's valuable marketing and sales intelligence. And sometimes it's indicative of personal interest. At other times it reveals content intent, not purchase intent. (h/t Andre Yee, Triblio's CEO who I first heard make this distinction.)
Intent data reflects different degrees of likelihood that specific action is indicative of actual research. It could be for education, entertainment, or any of a host of other reasons.
And inferred research in turn represents varying likelihood of purchase intent. It could represent job searches, competitive research, or other factors.
How do you know what intent data is really telling you? The best approach is an analysis based on a framework that aggregates data and signals.
Indications of Research in Context...That's Cool
It's hard to discern signal from noise in general, and purchase signal from background signal in any specific case. So how can you differentiate purchase intent and research from general activity?
Let's start by assuming the signal is indicative of relevant activity (vs. simply having an ad flashed before them for instance.) Once we're comfortable that the signals themselves are legitimate, then it's helpful to look at context.
And the best context framework is the combination of account fit and activity, and contact fit and activity.
A framework like this lets you see the intent data "forest" and still use the specific detail of the signal "trees."
For instance, significant aggregate account-level activity (certain number of users, certain percentage of typical buying team job functions, constellation of specific competitor and key term details, critical contextual™ technographic data, etc.) can help to differentiate non-research or purchase-related activity from significant projects. Then the specifics of contact level™ intent data signals (job titles, details/frequency/repetitiveness of each person's activity, stage in buying journey, role on buying team, etc.) help to inform the inferences which fill in the full picture.
While any single intent data signal (or small set) - whether an account level increase in activity for a topic at some point, or a specific contact that satisfies your key buyer persona - can be indicative of insignificant activity, in aggregate these account and contact activity clusters are much more informative.
Bottom Line
Buyer research insights aren't a simple product of intent data.
Intent data signals can indicate many different circumstances.
Depending on the collection method they may be essentially meaningless. In other cases, they'll be quite explicit.
Even when explicit, they can illustrate activity that is innocuous from a sales perspective. In other words, it might be simple engagement rather than evidence of deliberate research.
When intent data does indicate research, it's not necessarily indicative of purchase intent. It might be someone educating themselves for professional development rather than collecting information in support of a purchase.
And finally, even when it does indicate purchase intent, the actual intent may be different than salespeople assume.
That's why best practice involves a full marketing data stack, and data orchestration to help discern signal from noise, and purchase intent signal from general signal.
