Over the last few months, I’ve had a lot of clients bring up the topic of intent data. It’s gotten a lot of publicity and vendors are continually gathering new intent signals and creating new intent models, often through artificial intelligence techniques. But as we get into Hype Cycle season, I thought it would be worth trying to separate the hype from the reality.
I’ll state upfront that I’m a big fan of intent data and intent models. Companies that provide intent data and models have been prominently featured in our Cool Vendor lists over the last three years. And over that time, the value of the data and the models has steadily increased. And I’ve talked to plenty of TSPs that find intent enormously helpful when it comes to scoring leads, driving prospecting and demand generation activities or as the basis for account selection for an ABM program.
If you are selling into a well-established market, with a lot of activity/engagement related to that category taking place via content syndication sites, advertising networks, search and social media, intent data can be incredibly valuable to you, either on a standalone basis or as a major signal in predictive models. And IF there are enough companies that are showing high levels of intent to keep your sales teams busy and build a full pipeline, you should absolutely be taking advantage of it. After all, targeting accounts and individuals that show meaningful intent should yield a higher win rate and increased sales velocity relative to simply targeting companies that match your ideal customer profile or have high fit scores. What salesperson wouldn’t rather talk to someone in market as opposed to having educate someone who isn’t actively looking?
But that “IF” concept is incredibly important. There are a large number of scenarios where intent data and models don’t add nearly as much value (if any). It’s not because the intent data is inaccurate. It’s because there is simply not enough data available to use directly or to put in models. They include:
- New and emerging technology categories
- Certain geographies, industries or other niches
- Non-technology products
- Solutions (especially services) that can’t be easily categorized
Intent data is often readily available for technology infrastructure solutions, both hardware and software. And surge data can help predict when someone is likely thinking about replacing their ERP system. But I work with a lot of clients that don’t see value from intent data. Some of them are selling emerging technology and there aren’t enough companies actively in market. Other clients don’t have their categories included in the standard intent taxonomies or find that the searches and content consumption aren’t granular enough to be of much value. And others find that the intent data drives a “false positive” where the company is interested in a concept, but is merely trying to educate themselves around a topic that is getting a lot of buzz.
But there are also other issues. For some clients, intent data is inherently less valuable to them than technographic data. If you are going to run a competitive takeout campaign or try to attach a solution on top of a specific solution, the technographic data is the critical element, not the intent data.
And when it comes to predictive models, intent should be viewed as one of several key signals. There can be a tendency to overweight the intent signals in model building. If a predictive model yields a high intent score, but a low fit score, propensity to buy something may be high, but the propensity to buy your specific solution is going to be low. I’ve talked to several TSPs that find great value in intent data and models, but when it comes to lead or account prioritization and routing rules, the high intent/low fit ones get slotted behind both high fit/high intent and high fit/moderate or low fit. So the lesson is not to ignore the other fundamentals and rely too heavily on intent when there isn’t a good fit.
In summary, my advice is to leverage intent data where and when it makes sense. If there is enough raw intent data available, definitely consider acquiring it. And if you are looking at a predictive B2B marketing analytics solution, make sure you understand how useful intent data will actually be as a key driver of the model. And make sure you think about it across product lines, geographies and use cases. Some of the vendors may have stronger intent models, while others will have stronger fit models. One approach may make sense for acquiring net new clients for a solution in a particular geography, but make no sense for other geographies or for renewal and upsell/cross-sell models. Understand these considerations as you shortlist vendors and when you run proof-of-concepts.
Read Complimentary Relevant Research
How to Create a Data Strategy for Machine Learning-Powered Artificial Intelligence
MLpAI can help deliver systems with more automation and less human intervention, but success requires a data strategy to deal with the...
View Relevant Webinars
Big Data Architectures: Comparing Relational and NoSQL Databases
In the big data arena, few choices are more important and impactful than the persistent data store. Relational and nonrelational databases...
Comments or opinions expressed on this blog are those of the individual contributors only, and do not necessarily represent the views of Gartner, Inc. or its management. Readers may copy and redistribute blog postings on other blogs, or otherwise for private, non-commercial or journalistic purposes, with attribution to Gartner. This content may not be used for any other purposes in any other formats or media. The content on this blog is provided on an "as-is" basis. Gartner shall not be liable for any damages whatsoever arising out of the content or use of this blog.