To test new products, most companies rely on creating “minimum viable products” and testing customer feedback, or conducting focus groups or marketing surveys. There’s another method companies should try: “heat-testing,” or testing consumer reaction to online advertisements. Heat-testing is revolutionary because it takes place in the real world.
Since the invention of surveys in the 1930s, companies have used market research to size up consumer needs. Focus groups generate inputs about behavior and attitudes. Tools like conjoint analysis explore trade-offs consumers make when considering a product purchase. Large panels estimate the market opportunity for a particular product.
What do all of those market research methods have in common?
None of them take place in a real-life environment. In every case, consumers are aware that they are part of a research effort. What does that mean for a product marketer? They get a wealth of data about how people describe their behavior and attitudes. What’s missing? Data and insights about actual behavior.
There is an easily accessible way to solve this problem: online advertising as market research.
Responses to a digital ad — clicks, likes, email sign-ups — are more reliable indicators of purchase intent because they reflect how consumers behave when nobody’s watching. Testing via advertising captures real-life data about how customers respond to new product concepts, rebrands, and other big strategic moves. We call this type of research “heat-testing” — finding that spark between offering and audience that is the genesis of product-market fit.
Here are three solutions heat-testing offers marketers:
1. Validating demand for new product concepts
Let’s say you’re working on a new beverage concept: a dissolvable tablet in exotic flavors like elderflower and hibiscus that enhance a glass of water. The product offers sustainability benefits — not much packaging, no shipping of water — and the artisanal approach to flavors makes it feel a bit luxurious. It’s also customizable: combining flavors creates a personalized concoction.
There’s one question everyone wants to answer before they launch a new product like this one: Is someone going to buy this thing?
Conventional market research has numerous shortcomings in answering this question: Loudmouths in focus groups drown out important opinions, conditional questions abound in surveys (“If _ existed, would you buy it?”), and the extensive time and labor required for ethnographies and the like are not conducive to the pace of innovation that is required to stay relevant.
Fortunately, ad platforms lend themselves to multivariate testing, making it straightforward to access data-backed insights quickly. Platforms like Facebook, Instagram, Twitter, LinkedIn, and Google have instant access to millions of consumers who can be segmented into discrete audiences and targeted with multiple ads featuring a new product. Some ads will work and some won’t, providing data about where to retool variables.
Test variables may include a new product’s definition and features; the positioning, branding, messaging, and creative style that bring it to life; target consumer groups and calls to action to entice them; and other factors like pricing that make up the elements of a marketing campaign.
In the case of the beverage product, you might test three positions — Sustainability, Luxury, and Personalized Party — each brought to life by a couple of ads. (Ad creative alone should never be the reason a position succeeds or fails; using at least two creative approaches minimizes false positives.) And since your company wants to develop a younger customer base, you might target three discrete audiences ages 25–34 — let’s call them Planet People, Holistic Hipsters, and Spirits-Free Spirits — with ads.
Just like a regular marketing campaign, ads, landing pages, and supporting content are combined into a cohesive user experience. But a few elements are not like a regular marketing campaign: a) The presence of “in development” or “coming soon” to reflect the status of the product concept, b) the lack of pre-existing customer data, and, of course, c) the fact that there are multiple marketing strategies being tested in parallel.
How do you know if there is demand for your beverage concept? When a potential customer gives your company their email address for a product that does not yet exist, you’re on to something. In an era in which privacy is increasingly valued, email is currency; a high rate of email sign-ups is the best possible validation of demand without actually building the product.
2. Finding new customers for an existing product
Heat-testing can also identify growth opportunities for existing products. Repositioning a legacy product for new audiences is risky: Fresh messaging or imagery can alienate an existing customer base.
Testing new positioning on a small scale with new audiences identifies product-market fit, giving brands the opportunity to unlock additional revenue streams without significant investment. Additional testing of winning positions with core audiences — again on a small scale — can pinpoint any risks presented by new messaging.
3. Learning how target audiences respond to offers
The goal of heat-testing is to find the heat — the match between product and customer — and the marketing elements to create that match. Heat-testing is not A|B testing, which takes place in environments where traffic already exists and where users are automatically directed to variations of existing web pages to guarantee randomized sampling.
Testing strategy via heat-testing is different. Because the product and/or audience is new, traffic to landing pages needs to be generated, often in places where a brand lacks an existing mechanism to attract potential customers.
Generating that traffic for new product concepts isn’t a chore — it’s an opportunity to learn. Will an audience of “spirits-free” Millennials respond more to the Personalized Party position or the Sustainability position? Which combination of positioning, creative style, and audience works best to generate sign-ups for each concept? What proportion of people signed up to learn more once they reached the landing page? What does each email sign-up cost, and what can we infer about customer acquisition cost? Does behavior vary by ad platform? Answering these big, important questions de-risks expensive new marketing initiatives.
. . .
Heat-testing is revolutionary because it takes place in the real world. Every click on an ad is data. Every save or like is data. Every landing page view — yep, data. And when you compare the performance of every campaign variable, you learn a lot about how to position a new product. Best of all, those learnings are statistically valid.
Our approach to innovation takes Eric Ries’s minimum viable product (MVP) concept to a whole new level. In his book The Lean Startup, Ries espouses the benefits of launching an MVP and improving the product post-launch based on real user feedback. Heat-testing means that the MVP can be even more minimal than Ries imagined — that is, not even a product at all, but still something consumers can respond to.
Finding product-market fit via heat-testing requires iterating on lots of variables, so adhering to scientific principles is critical for a valid result. But capturing insights based on real-world behavioral data about purchasing decisions makes it worth the effort. The final product? A “heat map” showing the most responsive audience segments and the most efficient ways of engaging them.
Copyright 2022 Harvard Business School Publishing Corporation. Distributed by The New York Times Syndicate.