Super Seven Stars runs automotive dealerships, supermarket chains and industrial distribution across Myanmar. Before its content was rebuilt, the copy on its own website filed it under hospitality and tourism. Nobody inside the company would have described it that way, yet the page said so in plain sentences, and anything that read the page took it at its word. That is the problem entity SEO exists to solve, and it is a bigger problem for a company nobody has heard of than for one everybody has. A famous brand is described by thousands of other pages, so a wrong sentence on its own site gets outvoted. An unknown company has almost no other voices. Its own pages are the whole record, which means every classification error on them is the final answer.

The question that gets asked first

Most companies in this position ask how to rank for their own name. It sounds sensible and it points the effort at the wrong end. Ranking happens after a search engine or an answer engine has decided what you are, so if it has decided you are a tourism business, ranking first for your name just puts the wrong description at the top of the page. The better question is what a system would conclude about you if your site were the only thing it ever read. That means the conclusion drawn from the words that are actually there, which may have little to do with the one you intended. A search engine builds it by matching the name on your site to something it already recognises: a company, in an industry, run by particular people, doing particular work. When the page gives it nothing consistent to match against, it guesses or leaves you out.

Say what you are once, then say it everywhere

The fix has less mystery than the vocabulary suggests. State the classification plainly in the first lines of the page a stranger would land on, and use the same words wherever the company is described. Then add the machine-readable version: Google's documentation for Organization structured data lists properties such as the legal name, the URL, a logo and a sameAs list that points to other profiles of the same organisation, which is how you tell a system that the company on your site and the company in a business directory are one thing. After that, check the places that describe you without asking you, because that is where the contradictions usually live. Four of them are worth the first hour.

Where	What to check
Your About page	Does the first sentence name the industry the way a customer would
Social and directory profiles	Is the category the same one your site uses
The page listing your business units	Does each unit have its own description, or only a logo
Structured data on the home page	Do the type and name match the visible text

None of this needs a tool. It needs someone to read the company's own pages as a stranger would, which is rare, because the people closest to a site are the ones least able to see what it fails to say. The structured data is the part people expect to do the work alone, and it cannot. Markup that says one thing over a page that says another does not settle the disagreement; it writes it down in a format a machine reads. The visible words come first, and the markup repeats them.

Who has actually done this

Super Seven Stars is a Myanmar conglomerate spanning automotive, retail, wholesale, service, sales and distribution. The revamp began with the classification, replacing the hospitality and tourism filing with what the business actually does, and then rebuilt the two pages that carry the explanation. About Us was rebuilt around the Chairman's story. U Soe Lwin spent nine years in Japan working with established organisations such as Toyota before returning to Myanmar in 2003, and that history explains why the company operates as it does. What makes it useful to a machine and a person alike is that it can be checked. A founder described as a visionary leader gives a system nothing to attach to. A name, a country, a span of years and an employer give it four things to corroborate against other sources, and give a reader four things to verify for themselves. Businesses was the larger job. The page held a dozen-plus logos under one generic paragraph, which tells a reader nothing about any single business and tells a system that the company does something vague in many places. Each unit got its own accurate description, covering the dealerships, the retail and wholesale chains, the industrial distribution and the shipping. A design guideline went to the in-house designer alongside, so the page would read as one company rather than a logo wall. I should be straight about what that project does not show. I have no ranking data from it and will not invent any. What it shows is an order of operations: classification first, the founder story second, the business units third, all of it in plain words that agree with each other.

Your turn

This takes about thirty minutes and a text file. Write one sentence that says what your company is, in the words a customer would use to describe you to a friend. Then open your home page, your About page, your LinkedIn page and one directory listing, and copy out the sentence each of them uses for the same job. Put the five side by side. Expect at least one mismatch. Profiles get created in a hurry, by whoever had the login that week, and the category dropdown on a directory rarely offers the exact word your site uses, so a close neighbour gets picked and never revisited. Those small drifts are harmless one at a time. Together they are the reason a system cannot tell whether the company on your site and the company on a listing are the same one. Wherever two of them disagree about your industry, fix the wrong one before you touch any markup. Then search your company name in a private browser window and read what the first result says about you, because that text is all a stranger has to work with.