SEO case study · Content Software Product

Growing a new SaaS in search, and learning what scale costs.

Content Software Product is a SaaS that writes, optimizes and publishes SEO content for small businesses.

New accounts, 6 Jul – 5 Oct 2026
511
all channels, product analytics
Best week for signups
49
week of 14 Sep 2026
Google clicks, Sep 2026
2,489
first data Mar 2026: 15 clicks
Avg. position, Sep 2026
16.0
from 45.4 in Apr 2026

Client and project identities are withheld. Figures come from Google Search Console and project records. Organic leads and revenue are not disclosed.

The business

The business behind the search visibility.

It is our own product, so this case study shows our methods on our own site.

It competes in one of the most crowded search spaces there is: SEO and AI writing tools.

The challenge

Where growth was getting constrained.

Technical

  • Hreflang tags pointing Google at locale pages that did not exist.

Search

  • A brand-new domain in a crowded category.
  • Dozens of queries split across our own competing URLs.
  • Theory: AI-surface and fragment appearances were considered possible contributors to low blended CTR. The combined totals do not establish that cause.

Content

  • A large programmatic footprint where a few pages earned most clicks and thousands earned none.

Authority

Not assessed in our records

Website

Not assessed in our records

Conversion

Not assessed in our records

The strategy

The strategy was bigger than rankings.

  1. 01

    Publish where demand is

    Research-led content on AI writing and AI search that people actually cite.

  2. 02

    Shrink to grow

    Collapse thin pages, redirect what had value and retire the rest.

  3. 03

    Read the data correctly

    Segment queries and result context before changing titles. Blended impressions do not establish an AI-surface cause.

What we changed

What we actually built.

  1. Page-collapse migration

    27 Aug 2026: 4,361 thin page files removed; build time cut from about 11 to under 4 minutes.

  2. Redirect and retirement layer

    2 Sep 2026: 17,553 URLs 301-redirected and 4,485 returning 410.

  3. Title and description overrides

    24 Sep 2026: targeted fixes driven by Search Console, sitemap trimmed to 6,182 URLs.

  4. Multilingual programme

    Localized pages in five languages, started Aug 2026.

The data

What the search data shows.

Monthly Google clicksSource: Google Search Console, web search. Data begins 20 Mar 2026.
Monthly Google clicks
MonthClicks
Apr 26107
May 26304
Jun 26812
Jul 261,897
Aug 262,569
Sep 262,489

Top pages by clicks

  • AI writing benchmarks 2026510
  • Google AI Overview statistics331

90 days to 4 Oct 2026. Source: Google Search Console.

Outcome

What changed.

  • 511 new accounts across all channels signed up between 6 July and 5 October 2026. The best week, beginning 14 September, recorded 49 accounts in product analytics.
  • Monthly clicks grew from double digits to about 2,500 within six months of launch.
  • Average position improved from 45 to 16.
  • Original research pages became the strongest traffic drivers.
Read this alongside the numbers
  • This is our own software product, not a client.
  • Signups are counted across all channels from product analytics, which began tracking on 6 July 2026; they are not attributed to organic search alone. Weekly signups fell to 22–24 in late September.
  • Impressions are heavily concentrated in the US with low click-through.

Not disclosed: Revenue attributed to organic search; Signups before July 2026; Backlink and domain authority data.

What we learned

What this project taught us.

  • Blended CTR misleads

    Theory: a changing result-surface mix could affect blended CTR. Segmented evidence is needed before attributing this project's ratio to AI or fragment appearances.

  • Most "bad titles" weren't

    Of 55 near-miss pages, 42 had no title or description problem. Check the data before rewriting.

  • Scale has a cost

    Thousands of thin pages added maintenance and cannibalisation without traffic. Fewer, stronger pages won.

The system

Where this project sat in our method.

  1. Diagnose
  2. Architect
  3. Create
  4. Strengthen
  5. Convert
  6. Compound

Understand the terms used in this case study.

Organic traffic

Organic traffic is website traffic attributed to unpaid search results. A visit is a channel observation, not proof that the visitor became a qualified lead.

Attribution

Attribution is assigning credit for an outcome to marketing interactions under a defined model. Different models and data limits can produce different answers from the same customer journey.

SEO · Content · Local · Web Design

Want to build the same kind of system for your business?

Every business has a different search landscape. Let's identify yours.