Sample report

Read the whole thing before you sign up

Screenshots only get you so far. Below are four complete TrailMap reports, produced by the same pipeline a paying customer runs and exported by the same button they click. Nothing is gated, shortened, or dressed up for the web. Open one and read it.

These are two invented businesses

Peak Plumbing Co and Lumen Home Goods do not exist. Their accounts are generated from a fixed seed and cover March 1 to May 31, 2026, with problems deliberately planted in them: budget-limited campaigns, a CPA outlier, low quality score clusters, wasted search terms, and creative fatigue. We use synthetic accounts here rather than a real client's data, and the engine has no idea which findings were planted. Everything it reports, it found.

Pick an account

Two shapes of account, two views of each

The analyst view is the working document: every finding, the evidence behind it, and the recommendations ranked by what they are worth. The client view is the same run rendered for the person paying the invoice, which is what agencies on Growth and above send out under their own branding.

Peak Plumbing Co

Lead gen home services. About $43,000 across three months, roughly $14,000 a month, split over Google Ads, Microsoft Ads, Meta, and TikTok. 6,231 leads at a $6.94 blended cost per lead against a $45 target. The report raises 33 findings.

Analyst view → Client view

Lumen Home Goods

Direct to consumer ecommerce. About $168,000 across three months, roughly $56,000 a month, over Google Ads, Meta, and TikTok. $652,000 in revenue at a blended 3.89x return on ad spend against a 4.0x target. The report raises 31 findings.

Analyst view → Client view

Each file is a single self-contained HTML document with no network calls in it, which is what TrailMap's export button produces. Save one to your desktop, disconnect, and it still opens. It needs JavaScript and a browser from 2023 or later.

While you read

Three things worth checking

Every number is checked

The engine computes each figure and the narrative may only cite it. A number in the prose that the engine did not compute gets the whole response rejected and regenerated. Pick a figure out of the summary and go find it in the findings below it.

The hedges are real hedges

Where the data cannot support a conclusion, the report says so instead of rounding up to a recommendation. The plumbing account's social conversions are a good example: the engine flags that it is reading counted conversions as leads and tells you what that assumption rests on.

The analysis is computed, not generated

Impression share lost to budget, search-term waste, creative fatigue, the reallocation math: a rules-based engine computes all of it from the account data before a model is called at all. What you read is that analysis phrased and ranked, which is why the evidence tables underneath every claim have numbers in them.

Now run one on your own account.

14-day trial, 3 runs, no credit card. Upload an export or connect the platforms directly, and the first report looks like the ones above.

Start free → Read the whitepaper