Market intelligence and marketing benchmarks

What market intelligence covers.

Market intelligence is the organized gathering of outside information that helps a business make decisions. For a marketing team, that usually means three things.

Competitive intelligence looks at what other companies in your market are doing, such as the ads they run, the offers they lead with, the searches they buy and the pages they publish. Marketing benchmarks compare your numbers with those of similar businesses, including click-through rate, cost per click and conversion rate. Trend signals show what's changing around you, such as shifts in search demand, platform rules, new ad formats or category regulation.

This information only matters if someone uses it, so we start by asking which decision the work should support.

When it makes sense.

Market intelligence is useful when you're planning a budget or entering a new market, when your results have changed and you can't tell whether the cause is you or the whole category, or when leadership keeps asking how you compare.

It's less useful when you already have clear problems you can fix from your own data. If your tracking is broken or your landing page doesn't work, that comes first, and a comparison with other companies won't help. Very small or very specific categories can also be hard to research, because little public information exists. We'll tell you if the available sources are too thin to justify the hours.

Where the information comes from.

A lot of useful information is public, and we start there. Meta's Ad Library, the Google Ads Transparency Center and LinkedIn's Ad Library show ads that advertisers have run or are running. Those libraries show what was published, and they rarely show how much it cost or how well it worked.

Auction Insights in Google Ads shows who competes with you in your own auctions, and your search terms show the language people use. Google Trends and Search Console help us read changes in demand.

Paid research tools can add estimates of traffic and spending. They're useful for direction and rough comparison, and we label them as estimates because that's what they are. All of this work relies on information that's legitimately available. We don't pose as customers, and we don't use other companies' confidential material.

How far to trust a benchmark.

Third-party benchmarks vary a lot in quality, and we think you should know that before you use them. Many published figures come from vendors who sell tools or services. Their sample is usually their own customers, who may not look like you, and the report may be written to support what they sell. That doesn't make the numbers useless, but it does mean they need checking.

We ask a few plain questions about any benchmark. Who collected it and how large was the sample? Which countries and what period does it cover? Does it separate your category from others, and does it define a conversion the same way you do? A conversion in one report may be a purchase, while in another it may be a page view or a click on a phone number. Averages also hide a wide spread, because a figure that blends very large and very small advertisers may describe nobody in particular.

When sources disagree or a figure can't be traced back to anything, we say so and leave it out of the decision. Your own past results, compared over time, are often the more reliable yardstick.

Watching platform changes.

We track platform changes for our own media work, because they affect the accounts we run. New policies, retired features, changed defaults and new ad formats can all alter what a campaign should do.

For clients, we can turn that habit into a regular note on what changed, whether it affects your accounts and whether anyone needs to act. We read the platform's own documentation where we can, since secondhand coverage often gets the details wrong. As for what's coming next, nobody knows for certain, so we report signals as signals and keep predictions clearly marked as our opinion.

How we measure this work.

Intelligence work is hard to tie directly to revenue, and we won't pretend otherwise. The practical test is whether it changed or confirmed a real decision, such as a budget, a message or a market choice.

We also check our own work. When we flagged something as likely, we look back at whether it happened. When a benchmark turned out to be a poor fit, we note it and stop relying on that source.