Map search demand to the decisions customers actually make

Keyword lists sorted by volume hide what buyers are trying to decide. Build a decision map instead, using Google research on exploration and evaluation.

Most keyword lists are sorted by search volume. That tells you how many people type a phrase, but not what they are trying to decide.

A more useful approach is to organise search demand around the decisions your buyers make. It turns a long spreadsheet into a plan the business can understand.

What the research says about decisions

In 2020, Google’s Market Insights team in the UK published Decoding Decisions, a study of purchase behaviour carried out with the consultancy The Behavioural Architects.

The researchers watched shoppers as they searched and browsed. They describe people switching between two mental states, “exploration” and “evaluation”, in what they call the messy middle between a trigger and a purchase.

Exploration widens the set of options. Evaluation narrows it down. The same person may move back and forth many times before they buy.

Shortcuts shape the choice

The study then tested six behavioural biases in 310,000 simulated purchase scenarios. One was category heuristics: rules of thumb such as the memory in a laptop or the carats in a diamond.

In the simulation, category heuristics produced the largest or second-largest effect in 14 of 31 categories, according to the report.

That finding is useful for search planning. Many queries are people looking for exactly those shortcuts: the specification, comparison or proof point that lets them decide.

A caution: these were simulated scenarios with consumer categories, run by Google, which sells search advertising. The results show how preferences can shift, not how your market behaves.

Step 1: List the decisions, not the keywords

Before opening a keyword tool, write down the questions a buyer must answer before choosing you. Do this with sales and customer service, who hear them every day.

Hypothetical: a Dutch payroll software firm selling to mid-sized companies might list five decisions:

• Do we need new payroll software at all?
• Should we outsource payroll or keep it in-house?
• Which providers handle Dutch and German payroll rules?
• How much will it cost, and how long will switching take?
• Can we trust this provider with sensitive employee data?

Each decision is a cluster of searches. The list is illustrative, but most B2B purchases follow a similar pattern.

Step 2: Find the language people use

Now use the tools. Google’s Keyword Planner suggests keywords related to your products, services or website, and gives estimates of how often each is searched every month.

Treat those numbers as estimates. They are useful for comparing phrases, less useful as precise forecasts.

Your own Search Console Performance report shows the actual queries for which your site appeared in Google Search. Google’s help page explains that it reports clicks, impressions, click-through rate and average position.

Search Console is often more valuable than any keyword tool, because it shows real language from people who already found you.

Step 3: Assign each query to a decision

Put each query cluster under the decision it serves. For the payroll example:

• “what is payroll outsourcing” sits under the outsourcing decision.
• “payroll software Netherlands Germany” sits under the provider-fit decision.
• “payroll software pricing” and “switching payroll provider” sit under cost and effort.
• “payroll data security certification” sits under trust.

Some queries will not fit anywhere. That is useful: it either reveals a decision you missed or a query that does not matter to your business.

Step 4: Mark exploration and evaluation

Within each decision, note whether a query is exploring (broad, early, comparing types of solution) or evaluating (specific, comparing named providers or features).

Exploration queries often need guides and explanations. Evaluation queries need comparisons, pricing, proof and clear next steps.

This is our analysis rather than a Google framework, but it follows the study’s description of the two states.

Step 5: Check what each page offers

For every decision, check which page on your site answers it. Many companies find that the trust and cost decisions have no proper page, even though they block deals.

Google’s guidance on helpful, people-first content asks whether your content leaves readers feeling they need to search again elsewhere. That is a good test for each decision page.

Step 6: Share the map with the business

Present the result as a decision map, not a keyword list. Sales leaders, product managers and finance can read a map of buyer decisions.

It also shows where marketing is helping customers decide and where it is silent. Gaps are easier to fund when they are framed as missing answers to buyer questions.

Who should own the map

Give one person responsibility for the map, usually in content or search, but make sales a co-owner. Sales hears new objections first.

A short monthly check is enough: which new questions came up in calls, and do they appear in the map? If a question keeps recurring, it deserves a page.

Common mistakes

Sorting only by volume. High-volume, broad queries are often exploration. They may matter, but they are not always where decisions get made.

Ignoring small queries. A search used by only a handful of people a month may come from buyers about to sign. In B2B, low volume is normal.

Treating the map as finished. Decisions change as markets and products change. Review the map at least once a year.

The limits of this method

A decision map depends on the quality of your customer insight. If marketing guesses the decisions without talking to sales or customers, the map will be tidy and wrong.

Search also captures only part of the journey. Buyers talk to colleagues, read reviews and attend events. Use the map alongside, not instead of, other research.