Handover
Pack
Everything we built, what each piece is for, and the order to run it in now that it is yours.
The on-site fix is built and waiting to go live. Every correction the scan called for sits on a preview theme, verified and ready. Publishing it is the one action that starts everything else, because the models cannot re-read a site that has not changed yet.
What You Received
Eleven documents, all of them live on this hub and all of them in the ZIP. The hub is the version you should read from, because it stays current if anything changes. The ZIP is your copy to keep, and it works with no login and no dependency on us.
| Document | What it is for | Ref |
|---|---|---|
| AI Visibility Audit | What we found on your site and why it was holding you back | n/a |
| AI Visibility Baseline | What the models said about you before any of the work shipped | SCAN-01 |
| Where You Show Up | Who wins the answers you are absent from, and what they did to get there | SCAN-02 |
| Entity & Wikidata Strategy | Making your name mean your brand instead of the Hebrew word | WIK-01 |
| On-site Content Pack | The H1, FAQs, schema and robots work, built and awaiting publish | OCP-01 |
| Content & Editorial Strategy | What to publish, in what order, and the briefs to write from | CES-01 |
| Off-site Outreach & Backlinks | The listicles, Reddit threads and creators to earn mentions from | OOS-01 |
| Customer Review Strategy | The reasoning behind the review targets and the system | CRS-01 |
| Review Outreach Playbook | The seeding blast, the always-on flow, and the three templates | ROP-01 |
| 90-Day Operating Plan | The weekly rhythm that turns all of the above into visibility | OPS-01 |
| Technical Setup | Every account and access grant used, and who owns each one | n/a |
Start Here
One action unlocks the rest. The rebuilt store has to go live before the content and outreach are worth starting, because everything downstream gets read through whatever your site currently says.
- One. Publish the rebuilt store. Every correction the scan called for is already built and verified on a preview theme: the homepage brand definition, the ten FAQs with schema, the ring-size and care guides, the founder and organization markup, the AI-crawler permissions, the review markup, and the smart-ring compatibility data on all sixteen jackets. It is the single highest-leverage step, and it is ours to ship, not yours. We held it deliberately so the baseline scan could capture your true before-state first. That is done, so it is ready to go.
- Two. Start the 90 days at go-live. Publishing is what starts the clock, and it is also what starts the operating plan. Publishing does not change the answers overnight either. The crawlers have to come back, and that takes weeks rather than days. The week-6 scan is how we check they have.
- Three. Claim the entity. The Wikidata and entity work is what separates your brand from the Hebrew word over time. It is slow to take effect, which is exactly why it should start now rather than later.
- Four. Run the review seeding blast. Past customers first, jacket buyers before anyone else. Then switch on the always-on post-delivery flow so it keeps running without you.
- Five. Send the first outreach wave. The named listicle and Reddit targets in the off-site doc. This is the slowest-compounding workstream, so starting it in week one matters more than doing it perfectly.
Do not start the outreach before the rebuilt store is live. If you earn a mention while the old pages are still the ones being read, you have just created a fresh third-party source repeating the wrong facts, and those are much harder to correct than your own pages.
What Moves Which Number
Your baseline has four numbers worth watching. This is which piece of work is aimed at each one, so you can tell what is doing the lifting when the week-6 scan comes back.
| Baseline number | Today | What is aimed at it |
|---|---|---|
| Named answers with a wrong fact | 50% | The rebuilt store going live, then entity work |
| You appear on a category question | 9% | Content strategy, off-site outreach |
| Present on Google's AI surface | 0 / 29 | All three workstreams, slowest to move |
| Mentions that are positive or better | 80% | Review playbook, protect and extend |
Read the last row differently from the first three. The first three are gaps to close. The fourth is already an asset, and the job there is to keep it while the other numbers move, and to make sure the good opinion is attached to the current version of the brand.
Your Operating Rhythm
Three workstreams run in parallel, every week, for the full 90 days. None of them work alone. The full detail is in the 90-Day Operating Plan, this is the shape of it.
| Period | What happens | Checkpoint |
|---|---|---|
| Month 1 · weeks 1 to 4 | Rebuilt store published, spearhead content out, first pitches sent, seeding blast done | Foundation live |
| Month 2 · weeks 5 to 8 | Cluster depth under the spearheads, month-1 pitches worked, review floor crossed | Week 6 scan |
| Month 3 · weeks 9 to 12 | Remaining pieces filled, digital PR push, reviews driven to target | Week 12 scan |
Access And Accounts
The access model through this engagement was deliberately narrow. Read-only first, edits on an unpublished theme copy, publish last. At handover it all comes back to you.
| Account | Who owns it now | At handover |
|---|---|---|
| Shopify collaborator account | Node AI, via Partner dashboard | You revoke |
| Shopify custom app / Admin API token | Created by you, read only scopes | You delete |
| Google Search Console | Never granted, never used | Nothing to do |
| GA4 | Never granted, never used | Nothing to do |
| judge.me | Yours, on your billing | Stays yours |
| Shopify theme and all edits | Yours, on your store | Stays yours |
| This hub and the documents | Node AI hosting, yours to keep | Stays live |
Shopify. Settings, then Users and permissions, then Collaborators. Remove the Node AI collaborator. This also kills the custom app token.
That is the only one. We were never given Search Console or GA4 access, so there is nothing to remove there.
Do this whenever you are ready. Nothing on your store depends on our access continuing, and everything we built keeps working without it. If you want us to run the week-6 and week-12 scans, those need no access to your systems at all, they read the public web.
How We Measure Again
The baseline is only useful if the next reads are done identically. We pinned every variable so the movement you see later is real movement and not a change in method.
| Held constant | What it is |
|---|---|
| The questions | The same 40, word for word, across six product areas plus brand questions |
| The models | The same four pinned versions, not whatever is current at the time |
| The runs | Three per question per model, so a one-off answer cannot skew the read |
| Web access | On, in every run. A model answering from memory alone would read flat |
| The Google check | The same 21 keywords, same AI Overview and AI Mode surfaces |
We are not promising a specific recognition percentage or a ranking by a specific date. What we are promising is that the measurement is consistent, so you can see exactly which numbers moved, in which pocket, and by how much.
Open Items And Limits
Five things worth stating plainly, so nothing here reads as more settled than it is.
- The rebuilt store is still to be published, and that one is ours. It is built and verified on a preview theme. We held it on purpose so your baseline could capture a true before-state, and now that the baseline is locked it is ready to go live. We will confirm the date with you before we publish, and a few small inputs are still open on your side: your Instagram handle, YouTube and Facebook URLs for the brand markup, and whether the re-plating service is something you actually offer.
- The competitor set is our pick, not a verdict. We chose Buffr, STRUGA, The Space & Meteorite Co. and The Great Frog as sensible reference points. The models surfaced others on their own, which we kept. If you think a different brand is your real comparison, tell us and we will track it from the week-6 scan onward.
- The Google AI read is the free tier. It captures whether an AI box appeared and who it cited, not the deeper paid signal. Treat that section as directional.
- Writing and publishing the content is not in this scope. You have the briefs, the priority order and the outreach targets. Turning those into published pieces every week is the work the 90-day plan describes, and it is yours to run unless you want us retained for it.
- The review targets are a floor and a band, not a guarantee. They are what the category suggests you need to be credible, not a number we can promise to reach on your behalf.
Where Everything Lives
The hub at yoreh.nodeagency.ai is the current version of everything, and it stays up. The ZIP is the same documents as standalone files, so you keep a copy that does not depend on us hosting anything. The walkthrough is 45 minutes with the baseline scan open, going through what we found and what to do first.
Every document was written to be read without us in the room, but a handover pack that leaves you guessing has failed. If a brief does not make sense, if a schema change breaks something, or if you want a second read on whether a piece of content is working, email cristoforo@nodeagency.ai.
The week-6 and week-12 scans need nothing from you. We run them against the public web on the pinned models and send the comparison.