Featured · Techniques
Seven named methods, with the steps
Ideas tell you what to do and tips tell you how to do it well. Techniques are something else: repeatable procedures with steps, that produce an answer.
These are seven we use. Each takes between ten minutes and an afternoon, each produces something specific, and none requires a supplier.
The four diagnostic techniques
Each answers a question the firm cannot otherwise settle. Run them in this order the first time.
The seven-second test. Sit somebody unfamiliar in front of your home page, take it away after seven seconds, and ask what the firm does and who for. Whatever they can still tell you is your actual home page.
The stranger's journey. From a search on a phone, as somebody with the problem, all the way through to making contact. Note every point you hesitated. Ten minutes; it finds most conversion faults.
The grid measurement. Check your local visibility from a series of points around each office rather than from your own desk, because proximity means your own connection is the least representative location there is.
The source ledger. Record every enquiry for one month with where it came from, including the telephone ones and the ones somebody mentioned in a corridor. A month of this beats a year of analytics.
Together these four locate the constraint. Most firms have never run any of them.
The seven-second test is uncomfortable and worth repeating annually, because sites accumulate compromises and nobody notices from inside.
The source ledger is the one with the longest tail. Firms that keep it for a year stop arguing about which channel works, because the argument becomes a table.
Three techniques that build
These are not diagnostics but procedures that compound if repeated on a schedule.
The twenty-referrer list. Write down the twenty people best placed to meet your kind of matter before you do, by name. Contact two a month with something useful. Review the list annually.
The question ledger. Every time a client asks something you have been asked before, write it down. It becomes the content plan, the FAQ and the speaking topics without any further work.
The parallel test. When changing anything — a script, a service, a page — run the new version alongside the old for a fortnight and compare on the same number rather than switching and hoping.
Each takes minutes per week and produces most of its value after a year, which is why they are rarely sustained.
The parallel test in particular is what separates firms that know a change worked from firms that believe it did.
None of the three needs anybody's approval, which makes them available to an individual attorney as much as to a firm.
Running them without it becoming a project
The techniques are cheap; the failure mode is turning them into an initiative that needs scheduling.
Do the stranger's journey this week. It costs ten minutes and it will change what you do next.
Start the source ledger today, in a spreadsheet, imperfectly. Waiting for the right system is how it never starts.
Book the seven-second test with a colleague who has never seen the site properly.
Run the grid measurement once a quarter, not continuously.
Keep the question ledger somewhere everyone can add to in seconds — the friction is what kills it.
Use the parallel test for every change worth arguing about, and stop arguing.
Write the constraint you found in one sentence and put it where the partners will see it.
Then repeat the diagnostics annually rather than continuously, because they are for locating problems rather than monitoring them.
Seven procedures, none requiring budget or permission, most producing an answer the same week. The reason they work is not sophistication; it is that almost nobody runs them at all.
FAQs
What is the seven-second test?
Sit somebody unfamiliar in front of your home page, remove it after seven seconds, and ask what the firm does and who for. Whatever they can still tell you is your actual home page, as opposed to the one you think you have.
What is the stranger's journey?
Attempting your own enquiry from a search on a phone, as somebody with the problem, all the way to making contact — noting every hesitation. Ten minutes, and it locates most conversion faults.
Why measure local visibility on a grid?
Because proximity weighting means results differ by where the searcher stands, and your own office is the least representative point available. A single reported position describes your desk rather than your market.
What is the source ledger?
A record of every enquiry for a month with where it came from — including telephone calls and the ones somebody mentioned in passing. A month of it settles arguments that a year of analytics cannot.
What is a parallel test?
Running a new version of something — a script, a page, a process — alongside the old for a fortnight and comparing on the same number, rather than switching and assuming. It is what separates knowing a change worked from believing it.
End Note
None of these is clever and all of them are unused. They work because they replace opinion with an answer, usually within a week and usually for nothing.
Do the stranger's journey this week, start the source ledger today, and run the parallel test on the next change anybody disagrees about.
Key Takeaways
Techniques are procedures with steps, not suggestions.
Four diagnostics locate the constraint; most firms have run none of them.
The stranger's journey costs ten minutes and finds most conversion faults.
The source ledger turns channel arguments into a table.
Parallel test every change worth arguing about.
