How we get Speer found on Google and in AI answers

What wasn't working

In June an AI-assisted, unreviewed code change switched off Google indexing for the whole site. Nobody noticed for three months. Google removed our pages and search visits fell by two thirds.

What we fixed on September 30

• Turned indexing back on and added a guard so the same change cannot hide the live site again.
• Made every page readable by bots. The site used to send blank pages to Google and to the ChatGPT, Perplexity and Gemini crawlers. Every URL is now a real page with its own title and description.
• Fixed the metadata Google reads: canonical address tags, pages that returned errors, the sitemap, and a description file for AI crawlers.

What the October numbers show

• The recovery is real, but it is people who already know us. Almost nobody reaches us by searching for the services we offer. The plan is about changing that.
• For "speer technologies" we rank 6th, below LinkedIn and an old speer.ai test site. We are not first for our own company name.
• 72 pages are still waiting to be indexed. Re-indexing takes weeks. We can request it manually for the important pages.
• Most of our traffic is Canadian, so every page has to rank for Canadian searches as well as US ones.
• AI referrals: about 25 visits since June. Small, as expected at this stage.

Sources: Google Analytics (Jan 1 to Oct 29) and Search Console (Jul 30 to Oct 29), pulled October 29.

Why AI answers matter now

Recently a US healthcare startup asked ChatGPT who built the Baker Health app. It named Speer, and they contacted us. The deal is on hold for budget reasons, but that is the path we want to repeat.

When we tested ChatGPT, Gemini and Perplexity ourselves in September, all three credited Baker Health LLC and could not find Speer. That lead was a one-off. The plan is to make it happen consistently.

The plan: five pillars

• Done
  Fix the foundation

  Indexing, readable pages, sitemap. Still open: speer.ai, one address everywhere, Google reviews.

• In progress
  Industry pages that match what buyers type

  Eight pages, one buyer phrase each, partners named, short FAQ. Launching this week.

• Next
  Be the answer when someone asks about our work

  Case studies for Costco Pharmacy, Grata and Baker Health that state Speer built the product, confirmed by the partner on their own site.

• Next
  Earn mentions elsewhere

  Google reviews, a Clutch profile, "top AI developers in Toronto" lists, press on the 35 patents.

• Next
  Measure monthly

  Pages listed, visits from search, and how many of 20 fixed AI questions mention Speer.

Searches we want to show up for

Most of the apps we built get no searches by name, so we focus on the three partners people do search for, plus the services we sell.

"grata" alone gets 5,800 US / 590 CA but is shared with other companies of that name. "costco pharmacy" alone gets about 850,000 US, but those are people looking for Costco itself.

What buyers actually search

Pages need to rank in both countries. Canadian searches are usually "Toronto" plus the service. US searches are the industry plus "development company".

Same data and window as the table above. "AI in healthcare / finance / education" each get 10,000 to 12,000 US searches a month (750 to 3,000 in Canada), mostly from people researching rather than buying. An explainer on each industry page is what AI tools cite later.

The pages we will build

One page a week, in this order. Each is written around one search.

Volumes are the same Google Ads data as above, summed where a page targets more than one phrase. Industry pages are already built; the remaining work there is the FAQ block.

Every page follows the same template

• Title and heading use the search phrase, not a tagline
• First 100 words say what we do, for whom, with proof (partners, patents)
• Partners named, with a link to the case study where one exists
• Four to six FAQ questions in the words people actually search
• Compliance and location details where they matter (HIPAA, PIPEDA, Toronto, New York)
• One call to action: book a call with Keegan
• 800 to 1,500 words of real copy, not a hero image and logos
• Structured data so Google and AI tools can parse it

Competing in the US

US search volumes are 10 to 30 times Canada's for the services we sell, but today only 48 of 967 Google clicks in the last three months came from the US. Six things change that.

• Write for the US buyer. Industry and service pages use the US phrasing ("healthcare app development company"), with Toronto in the company pages only. Google decides who sees a page by its content, not by where our office is.
• Lead with US proof. Baker Health (New Jersey), Costco Pharmacy, State Farm, Atlantic Records and Carrier Global go first on the pages and in case studies, with the city or state named.
• Make the New York office real to Google. A Google Business Profile for 26 Broadway, the same address on LinkedIn and the site, and "serves Canada and the United States" in the page metadata.
• Get listed where US buyers look. Clutch and GoodFirms profiles with reviews, and the "top healthcare app development companies" lists that US buyers and AI tools both read.
• AI answers have no border. ChatGPT recommends the same firm to a buyer in Toronto or Texas. The case-study and review work is the US play as much as the Canadian one.
• Track it separately. US clicks in Search Console, reported monthly. 48 today; the target is 300 a month by Q2 2027.

Not on the list: buying links or ads for US rankings, or duplicate "USA" versions of pages. Google penalises the first and ignores the second.

How we expect it to grow

SEO builds up over time. Pages and mentions keep bringing traffic for years, but results take a few months to show. October recovered quickly because brand searches came back. Getting found by people who don't know us takes longer.

Risks: case studies that don't clearly say Speer built the product, no reviews collected, paid links left in place.

How social media plays a role

Search engines and AI tools don't see likes or engagement. They see names, links and reviews. Social media is how we get those.

• Reviews. Five to ten Google reviews from past partners in 60 days, then Clutch.
• Partner co-marketing. One public mention per partner: a line on their site, a post, or a quote in our case study.
• Every page gets a post. Name the partner and link to the page. The name and the link are what search tools pick up.
• Lists. Get Speer onto the "top AI developers in Toronto" lists our competitors are already on.
• Cadence. One page or case study a week.