Assessment
What your website, your search presence and your data are actually doing, checked against a written instrument rather than an opinion. Findings ranked by what each one costs you.
What it covers →
Services
We look at what your marketing is actually doing, tell you what we found, and then fix it. 25 years of running an internet business sits behind every answer.
Three things
They connect, but none of them requires the others. Most people start with the first and stop there.
What your website, your search presence and your data are actually doing, checked against a written instrument rather than an opinion. Findings ranked by what each one costs you.
What it covers →The channels audited, then run properly. Writing, posting, engagement and editorial, measured against a baseline set before anything changes.
What it covers →Amazon and Walmart. Catalog, feeds, listings, variations and the day-to-day work that catches most brands out after launch.
What it covers →Assessment
Almost every engagement starts here, because almost nobody knows what their marketing is doing. Not because they are careless. Because nobody has looked closely, and looking closely is a specific job.
We look at the site the way a search engine and an AI assistant see it, and write back with the 3 things we would fix first. No login, no meeting, no obligation, and nothing to pay.
Send an address →Scored against a written instrument and run entirely from public sources. Titles and metadata, structure, canonicals and redirects, structured data, page weight, security headers, orphan pages, retrievability by AI assistants.
A list of everything that could theoretically be improved is not useful. The report is ordered by what each finding costs the business, so the first page tells you what to do first.
Metadata, redirects, structured data, page speed, broken tracking. Most of it shows a difference within 2 weeks, and we tell you which items will not.
What that looks like →A crawl is precise about what is technically there and silent about everything else. It will not tell you whether your photography is working, whether your product naming helps or hinders, or whether one site can serve two different kinds of buyer. Where a question is outside what the evidence covers, the report says so rather than guessing.
Marketplaces
Selling on a marketplace is its own discipline — different data, different rules, different economics, and a day-to-day workload that catches most brands out after launch.
Every platform wants the same product described differently. Titles, attributes, categories, images, feeds. That is where the work is, and where the errors hide. A company needs six-figure revenue potential on the platform for marketplace work to be worth the time on either side. Most small businesses do not clear that, and we will say so.
Most companies come to marketplace work with no way to judge who can do it. There is no license and no register, and the credentials that exist cover advertising rather than catalog and account work. The ordinary way of choosing a supplier — shortlist three, compare them, pick one — does not produce a comparison here, because the things you would compare are mostly not published.
We surveyed the firms offering this service. Most had never been reviewed by anyone, and most had never been hired more than a handful of times. The one thing nearly all of them published was a price, and the price tracked nothing — the firms with no track record at all were charging many times more than the firms with one.
So we put the evidence on the table before you decide anything. You get a written assessment of your own catalog and your own market, built from data you can check, and it says plainly whether this channel is worth your money. Sometimes the answer is that it is not, and we will say so with the arithmetic attached.
Getting a product range onto a platform correctly the first time, structured so it can be maintained rather than rebuilt.
One master database feeding several platforms, each wanting the data in its own shape.
Titles, bullets, attributes and images against what people actually search for on that platform.
Size, color, style and price variants — among the hardest things to set up correctly and the easiest to break.
Start here
Three specific things worth knowing about, from public information only.
Social and content
Run properly, and measured.
Most social programs are busy rather than effective, and almost none are measured against anything. The work starts with an audit, then a baseline, then the program itself — so at the end of it there is an answer rather than an impression.
Audit first
Every channel examined as it stands: what is posted, how often, what is engaged with, what is dormant, and which channels are declared but never worked. Followers are the least interesting number on the page.
A baseline before anything changes
The numbers as they are today, recorded before the first post goes out. Without it, nobody can honestly say later whether the program worked.
Writing and editorial
Posts, captions, longer editorial and blog content, written to a plan rather than to fill a calendar. Built around what your audience actually searches for and reads.
Posting and engagement
Scheduling, publishing, replying and community management across the channels that matter for your business, rather than all of them because they exist.
Working with what you already have
Most companies are sitting on material that never reaches the page where the buying happens. Connecting what exists is usually a smaller job than commissioning something new.
Reported against the baseline
Monthly, against the numbers recorded at the start, including what else changed in the same period. Every agency reports the number that went up. Almost none records what else was happening.
The research behind this →Where the honest difficulty is
Attribution on social is genuinely hard, and anybody who tells you otherwise is selling something. What can be done is to record the starting position, record what changed, record what else was going on at the same time, and be straight about how much of the movement can fairly be claimed. That is a slower answer than a dashboard, and a more truthful one.