Starting position
What the site looked like when we opened it
We took the site over with search traffic already falling. It was not falling because of anything anyone did in June — the slide runs back through March, and the steepest drop happened in late April, before we had access.
Our opening audit scored the site 45 out of 100. The specific things we found:
| What we found | The number | Why it mattered |
|---|---|---|
| No Search Console access | — | Nobody could see what Google was doing. Verified 5 July. No manual penalty — confirmed clean. |
| Spam backlink injection | ~200 links | Four active campaigns pushing junk links at the domain. Site spam score 26. |
| A hidden link-exchange page | 5 outbound | Linking to five competing dispensaries. The exact pattern Google penalises for. |
| Plugin holding a former vendor's login | 1 | A previous SEO contractor's account was still authenticated inside the site. |
| Brand pages with no main heading | 66 pages | And four internal links between them in total. Google had nothing to follow. |
| Duplicate copy across the brand library | 1,447 words | The same Alien Labs text on every brand page — 64% of each page, identical. |
| Dead pages and demo content | 31 URLs | Tourist-attraction pages and theme placeholder posts, some lorem ipsum from 2018. |
| Broken ordering path | 1 | The Weedmaps menu was blocked from loading. Customers hit a dead end. |
Search performance
The numbers, plainly
Clicks are down over the quarter. That is the honest headline and we are not going to dress it up. What matters more is the shape of the decline, because that is what tells you whether the work is landing.
Three things to take from those two charts
The bleeding is slowing. Month-over-month clicks fell 27% into June, 13% into July, and 4% into August. Each month the drop is smaller than the one before it. That is the pattern you want to see before a recovery, and it started after the work started.
Impressions stopped falling. July to August was 128,591 to 127,994 — flat. That is the first month since March that Google did not show the site materially less often. Impressions turn before clicks do.
The traffic that remains is better traffic. Click-through rate went from 0.68% in March to 0.96% in August — a 41% improvement. Fewer impressions, but a much higher share of them turn into an actual visit. That is titles, headings and structure doing their job.
| Quarter over quarter | Mar–May | Jun–Aug | Change |
|---|---|---|---|
| Clicks | 6,724 | 3,979 | −40.8% |
| Impressions | 881,236 | 441,422 | −49.9% |
| Click-through rate | 0.76% | 0.90% | +18.4% |
March is a partial month in Search Console's window (from the 5th), so the quarter comparison uses what Google has. The month-by-month figures above use complete calendar months only.
Early signal
Every new city page is indexed and appearing
The eleven city pages plus the hub went live in August. All twelve are in Google and all twelve are already being shown in results — measured over the last 28 days. For pages that are weeks old, appearing at position 11 on their own city term is the right first step.
Clicks from these pages are near zero so far, and that is expected: nothing gets clicks from position 11. The job of the next 60 days is to move them from page two to page one.
| Page | Times shown | Average position | Status |
|---|---|---|---|
| Santa Ana — the flagship | 286 | 11.2 | Indexed |
| Areas We Serve hub | 89 | 5.9 | Indexed |
| Costa Mesa | 65 | 11.3 | First click |
| Westminster | 25 | 35.8 | Indexed |
| Anaheim | 24 | 32.0 | Indexed |
| Irvine | 24 | 16.2 | Indexed |
| Orange | 20 | 11.7 | Indexed |
| Newport Beach | 14 | 35.6 | Indexed |
| Huntington Beach, Tustin, Fountain Valley, Garden Grove | 2–5 each | 10–64 | Indexed |
Deliverables
What we built and fixed
Forty-five items across six workstreams. Everything below is live on evergreenoc.com unless it says otherwise.
Site structure and new pages
9 items- Built the Areas We Serve section — eleven city pages plus a hub Santa Ana, Anaheim, Garden Grove, Irvine, Tustin, Costa Mesa, Orange, Westminster, Fountain Valley, Huntington Beach, Newport Beach. The plan called for four. We built eleven.
- Santa Ana is the flagship The single highest-value page that did not exist. It carries "dispensary santa ana" — 1,900 searches a month at low competition, the term we have been chasing since day one.
- Restructured the site into real sections Brands, strains and service areas are now their own content types with their own sitemaps, instead of 200-plus loose pages. This is what makes a library this size manageable.
- Redirected every old URL to its new home The old delivery-areas and service-area addresses all point cleanly at the new city pages. Nothing that was ranking or linked got dropped in the move.
- Moved seven stray pages under the right section Brand and strain pages sitting loose at the top level, filed properly, with redirects so no link breaks.
- Consolidated a duplicate strain page Two pages were competing for Permanent Marker. The weaker one was retired and redirected into the stronger one, and the three internal links were repointed.
- Rebuilt the Order Online page The Weedmaps menu was blocked from loading, so customers were hitting a dead ordering path. Replaced with a working gateway straight to the full menu.
- Built and wired up the Brand Deals section Deals can now be updated without a developer. We documented the artwork rules the cards need, since the banners Weedmaps supplies do not fit the format.
- Cleaned up the navigation Three brand pages added under Brands, three duplicate top-level entries removed.
The brand library rebuild
5 items · 67 pages- Found and removed duplicate copy on every brand page All 66 brand pages were carrying the same three sections of Alien Labs text — 1,447 words, 64% of the page, identical everywhere. The Stiiizy page was asking customers "Does Evergreen carry Alien Labs?" in visible body copy. One shared template caused it, so it was one repair rather than sixty-six.
- Wrote original copy for the library and put it live Purpose-written copy per brand, built around the searches that brand actually gets. Verified live: Stiiizy and Wyld now carry roughly 1,500–1,650 unique words each, their own page title and their own heading. No shared text remains.
- Rebuilt the brand page template across all of them Every brand page now has a proper main heading, a clean heading structure, and alt text on 14 of 15 images. Before this they had no main heading at all.
- Rebuilt internal linking on the brand pages Each brand page now links to 12 related strain pages. Across the whole library there used to be four internal links in total. That is what starts moving authority through the site instead of dumping all of it on the homepage — and brand-to-brand and city links are the next stage of the same rebuild.
- Added main headings to the strain pages Same fix on the strain side. The heading order and image descriptions there still need a second pass — it is in the next-90-days plan.
Technical SEO
10 items- Rebuilt the business information Google reads The code identifying the business had no phone number, no hours, and junk data in it — including a stray reference that pointed at a Guyanese stew. It now carries the correct address, phone, 6am–10pm hours and the ten Orange County cities served.
- Verified Search Console and confirmed no Google penalty Worth saying plainly: nothing has been manually flagged against the site. We also captured a full traffic baseline, which is why every number in this report exists.
- Installed proper redirect management The site had no way to handle redirects at all. It now tracks URL changes automatically, so pages do not quietly turn into dead ends during a rebuild.
- Pulled 27 thin tag pages out of search Near-empty pages competing against the real product and strain pages for the same terms.
- Pulled the Uncategorised archive out of search 37 posts were sitting under a meaningless heading and being indexed that way.
- Pulled six empty category pages out of search Left over from the demo-content clear-out, with nothing on them.
- Retired 31 dead pages Ten leftover tourist-attraction pages and 21 theme demo posts. Nothing permanently deleted; everything recoverable.
- Rescued an old gallery URL It had 22 links pointing at it from other websites and would have started returning an error after the cleanup. Redirected, so that value stays with you.
- Hid the in-progress redesign from Google The draft homepage was competing with the live homepage in search results. Set to reveal on launch day, then swapped in.
- Fixed the "WECLOME TO EVERGREEN OC" typo It appeared 240 times across the site. All corrected.
Local search and reputation
4 items- Audited Google Business Profile and every directory listing Checked Yelp, Weedmaps, Leafly and Cannapages. Found a name mismatch across platforms, an unclaimed Leafly listing, and Leafly publishing 9am opening when you open at 6am.
- Locked the canonical name and address across every platform Every listing now uses identical wording. Inconsistent name and address details are one of the most common reasons a business does not surface in the local map results.
- Quantified the review gap 294 Google reviews at 4.5 stars, against 420 Central at 3,432, Tropicanna at 2,574, Eaze at 1,626, From The Earth at 1,547 and Mr. Nice Guy at 708. Reviews are the single biggest lever on getting into the map pack, and it is the largest gap we measured.
- Prepared a citation and press pack Directory submissions ready to go, plus draft pitches for Voice of OC, the OC Register and the OC Business Journal.
Security and cleanup
5 items- Removed a plugin exposing a previous vendor's login credentials Deactivated and deleted entirely, directory confirmed gone.
- Deleted a hidden link-exchange page It linked out to five competing dispensaries with matching map embeds. This is the kind of thing Google penalises sites for. Gone.
- Deactivated eight unused plugins Less to load, fewer ways in.
- Backed everything up before touching it Full database snapshot and a complete URL inventory taken before the cleanup, so every change is reversible.
- Organised the media library All 617 images sorted into seven folders so your team can find things. No files moved on disk, so no image on the site broke.
Research, strategy and content
12 items- Full SEO audit at the start of the engagement Scored the site 45 out of 100 and produced the priority list nearly all of the work above came from.
- Index coverage audit across all 162 brand and strain pages Checked every page individually. 80 of 162 were actually in Google, and the ones being skipped were consistently the ones with the most search demand behind them — of 21 pages targeting big terms like Cookies, Stiiizy, Blue Dream and Wyld, only five made it in. That single finding reframed the whole content plan.
- Built and validated the internal linking map Mapped all 346 links inside page content. 159 pointed straight back at the homepage, only 14 pointed at a brand page, and 117 pages had nothing linking to them at all. This became the blueprint for the linking rebuild.
- Duplicate content investigation Tested whether copied or thin writing was causing the indexing problem, which is how we found the shared Alien Labs block.
- Backlink audit, monitoring and disavow file 846 backlinks across 257 domains, with four active spam-injection campaigns. Roughly 150 junk domains from a previous vendor's link building are written into a disavow file, ready to submit. Monthly monitoring is set up to catch the next injection.
- Competitor keyword gap analysis 200 search terms your competitors rank for and you do not, with volume and difficulty attached, so we pick off the winnable ones first rather than guessing.
- Wrote a 90-day local search strategy A full plan for owning Santa Ana and Orange County searches, built on live search data: page-by-page changes, the new location pages, a six-post blog plan and a month-by-month roadmap with targets.
- Broke the strategy into a 40-task register Every task tracked individually, which is why this report can count them.
- Wrote the first competitor interception article Targeting a high-intent local search worth 3,600 searches a month at almost no competition, where a rival dispensary currently ranks with a much thinner post. The article and its structured data are written and ready to publish.
- Produced the launch migration runbook Nine sequential phases with the exact click path and a verification step for each, so the redesign launch did not lose rankings. Nothing counted as done until it had been checked.
- Built the deal card production system A repeatable build for Brand Deals artwork, with the cutout and layout work automated so new deals do not need a designer each time.
- Wrote and shipped the Zikigai Garden × Elusive drop campaign Announcement and live-now emails to Green Club members, with the compliance language and the ordering cut-off times built in.
Verified 1 September 2026
The state of the site today
We re-crawled the live site the morning this report was written. These are the actual numbers, not the plan.
| Section | URLs live | Notes |
|---|---|---|
| Brands | 67 | Own content type, own sitemap, unique copy per page |
| Strains | 98 | Own content type, own sitemap |
| Areas We Serve | 12 | All indexed, all appearing in results |
| Posts | 33 | Down from 54 after the demo-content clear-out |
| Pages | 22 | Trimmed to 8 on 2 September — see the findings below |
| Categories | 3 | Down from a sprawl of thin archives |
| Total in sitemap | 235 | One clean sitemap index, correctly linked from robots.txt. Now 221 after the 2 September trim. |
Technical findings
Things we are updating next
Three items came out of the site audit. Two have been completed since this report was written; the third is the priority objective for the coming quarter. None of them need anything from you.
Connect the product library to the city pages
The twelve city pages are live, indexed and already appearing in Google at positions 11 to 36. Moving them onto page one is the priority objective for the coming quarter, and internal linking is what does it — Google follows links to work out which pages matter.
Two pieces are already in: the navigation now points straight at the city hub rather than through a redirect, and the footer links every page on the site to all twelve cities. The next stage is connecting them in context — a Stiiizy page mentioning same-day delivery to Irvine, a strain page pointing at the cities it serves. That is two template changes covering all 165 brand and strain pages at once.
Streamline the sitemap
The pages sitemap still listed 13 old /delivery-areas/ and
/service-area/ addresses left over from the restructure. Google was spending its
crawl allowance following those instead of reading the new city pages.
Done on 2 September. Every old address still redirects correctly, so nothing that was linked or bookmarked has broken, and the sitemap has been resubmitted to Google. With the About page consolidated below, the pages sitemap is now 8 clean addresses instead of 22 and the site total has gone from 235 to 221.
Consolidate the two About pages
Two versions of the About page were live, competing with each other for the same search.
Done on 2 September. /about/ now redirects into /about-us/, which
carries the rebuilt copy, so there is one About page and it is the good one. The pages sitemap is
down to 8 clean addresses and the site total to 221.
Waiting on you
One thing we need from your side
- Claim the Leafly listing Leafly is unclaimed, and it is currently telling people you open at 9am when you open at 6am. Claiming it has to come from your side — we cannot do it on your behalf. Once it is claimed, hours, menu and profile details are ours to keep current along with everything else. UNBLOCKS — corrects the wrong opening hours and completes the listing set
Everything else on the list — the review request programme, the Google Business Profile updates, the old admin accounts and the disavow file — is ours to run. We are handling those internally and will report on them as they land.
The next ninety days
Where the work goes from here
The next quarter is about finishing the storefront, connecting the site together, and turning on the channels that bring people back. Sequenced because several pieces depend on the ones before them.
- Complete the Meadow store. The Shop page is built and the menu runs on Meadow, your point of sale. Finishing it means getting every product image downloaded, sized and matched to the right item so the catalogue reads correctly. This is the biggest single piece of the month and needs a team member on the build alongside us.
- Rotate the correct product images onto the brand pages. Each brand needs its own current products shown, not placeholders. This runs alongside the store work because both draw on the same image library.
- Connect the product library to the city pages. In-content links from brand and strain pages to the cities they serve — two template changes covering all 165 pages at once. This is the change most likely to move the city pages onto page one.
- Publish four articles this month. A four-a-month cadence written specifically to support the service area pages — each piece targets a term the keyword research shows is winnable and links back to the city page it belongs to, so the articles feed the local pages rather than sitting on their own.
- Keep updating the strain and brand copy. The library is written; this is the pass that brings heading structure, image descriptions and internal links up to the same standard.
- New store photography. A fresh shoot this month, feeding both the Google Business Profile photo count and the brand pages.
- More local citations. Around fifteen were built and indexed last month; we continue working through the list.
- Green Club email campaigns. Ongoing sends to the member list, in the same format as the Zikigai drop.
- Green Club text campaigns. The safety mechanism is built — a verification gate that confirms consent before anything sends, so the programme stays inside cannabis SMS rules.
- Double down on reviews and reputation. Already running; this month we scale the effort with the team behind it, because reviews remain the strongest lever on local ranking.
- Launch the voice agent. Now the site is live it can go on. It will not verify product availability until the Meadow catalogue is finished and the brand and strain pages are current — we would rather it say nothing than quote a product you do not have on the shelf.
- Begin foundational backlinks. Groundwork starts this month; the first placements land here, building the domain authority the city and brand pages need to climb.
- Build out the delivery pages. Rebuilt properly as delivery rolls out, so each area has a real page rather than a redirect.
- Watch indexing and report the number. With the templates, the linking and twelve city pages all live, this is the window where we find out how much of the library Google picks up. 80 of 162 is the baseline to beat.
- Article cadence continues at four a month.
- Scale backlinks and citations. Both compound — the earlier they start the more they are worth by the end of the quarter.
- Build the brand pages the keyword gap identified. Terms your competitors win purely because their menu pages are crawlable — 710 Labs, Raw Garden, Plug n Play, Boutiq and the rest, for the brands you actually carry.
- Extend the city pages to Fullerton and the other towns where competitors rank with nothing more than their homepage.
- Local press outreach. Voice of OC, the OC Register and the OC Business Journal — pitches are drafted and waiting on a spokesperson and a few company facts.
What the sequence depends on
Finishing the Meadow catalogue is the gate. Until products are visible and correct on the site, the voice agent cannot answer questions about stock, and the brand pages cannot show the right products. Almost everything else in the quarter runs in parallel — but that one comes first.
The other dependency is people. The store build, the photography and the review programme all need a team member working alongside us rather than us working alone.