Bringing into focus
We pick the few queries that pay, and leave the rest blurred. One page, one question.
The method
Four stages, in this order.
Chiaroscuro
The audit points to the pages that matter, not the tool's four hundred alerts.
Leading lines
Architecture, content and links all point to those pages. Google understands where to look.
The Kanizsa triangle
AI describes you based on what others say about you. We make sure they say it.
Of the recommendations on this page, how many were tried out on a client's site first?
The lab
200 sites, a laboratory method.
With Newp, I track more than 200 sites day to day. It's a testing ground few consultants have: every idea is tested there like an experiment before it reaches a client.
- A hypothesis. Just one, worded so it can be proven wrong.
- A small sample. About ten sites get the change. All the others serve as controls.
- A measurement. Clicks, rankings, AI citations, over several weeks.
- A verdict. The test fails: we drop it and move on to the next. It succeeds: we roll it out to the relevant sites, not the others.
Test 1 / 4Hypothesis
Sample testedTest rejectedTest validatedRolled out
Lab notebook · click to replay a test
Fig. 1. Four tests, fast-forwarded. (Illustrative scenarios: each square is a site, real figures vary by test.)
The safeguards
Four things I don't sell.
- Guaranteed first pageNobody can honestly guarantee it. I commit to a plan, priorities and measurements.
- Twelve-month contractsSupport is monthly and can be cancelled. You stay because it works.
- A junior behind a salespersonThe person who answers you is the one who builds the strategy. That's me.
- A forty-page reportYou get what needs doing, in order, and what it achieved.
No clicking allowed.
Fig. 2. Non-exhaustive list. (The competition offers others.)
30 minutes · video call · with me
Be the one they remember.
Together we look at where you stand on Google and in AI, and what should change first.
Book a callFig. 3. A K. (The landscape was missing one.)