What we changed on our own practice’s website, and why.

The full story

A practice that was being seen but not chosen, and the specific work that followed.

The starting point

Over the three months to October 24, 2026, familyhopehealthcare.com appeared in Google results about 79,000 times and received 868 clicks, a click-through rate of about 1.1%. Its average position was 14.3: the bottom of the first page or the top of the second, where few people click.

The indexing report looked worse than it was. Google listed 98 addresses as not indexed, but 76 of them were old pages that no longer existed, and only five were pages Google had chosen not to show.

Titles that matched the search

Several page titles used our words instead of the patient’s. Titles such as “Depression Psychiatric Care in Knoxville” became “Depression Treatment in Knoxville”, and an ADHD page titled for an “evaluation” now says “treatment”, the word people were actually typing.

Five titles changed in all. Each change was made in every place the title appears: the page title, the sharing tags and the structured data.

Whole answers on narrow pages

Some city pages answered only one corner of a search. The Franklin depression page was almost entirely about pregnancy and postpartum care; two ADHD pages spoke only to adults even though children as young as four are seen there. Each gained a short, factual section covering the rest of the question.

Pages for “psychiatrist” searches in two smaller cities now say plainly that visits are with psychiatric nurse practitioners and that the physician Medical Director is not on the appointment schedule.

Links between related pages

The OCD service page did not link to the Gallatin page that was appearing for “ocd treatment gallatin”, and the Franklin city page was outranking the dedicated Franklin depression page it never linked to. Both now link where they should, alongside links between nearby cities’ pages.

Insurance pages that do not repeat

The practice has a page for each of four accepted insurance plans in each of eight Tennessee cities. We rewrote all thirty-two until not one paragraph repeated eight consecutive words found anywhere else on the site, and trimmed titles and descriptions to fit search results.

Dead addresses forwarded

Fifty-eight retired addresses were still returning “not found”: old city pages, old file names and towns the practice no longer has pages for. Each now forwards permanently to the closest live page, and all fifty-eight were tested after the change.

A weekly guide

The practice now publishes one plain-language guide each Monday, beginning with a walk-through of a young child’s first ADHD appointment. Each guide links to the pages where a family can request care.

How we chose where to start

We did not begin by writing new pages. We began with the Search Console performance report, sorted by page and then by search, and looked for one pattern: searches where the practice was already being shown often but sat between positions eight and twenty. Those are the near misses, pages Google already considers relevant but places just below where people click.

For each near miss we opened the page and asked one question: does this page fully answer what the person typed? Where the answer was no, the fix was usually small and specific: a title in the wrong words, a section that covered only part of the question, or a missing link from a page that already ranked.

Reading the indexing report without panic

A long list of “not indexed” addresses looks alarming, and many practices respond by asking Google to index everything. We sorted the list first. Most entries were old addresses that no longer existed and needed forwarding, not indexing. Only a handful were live pages Google had chosen not to show, and those were the ones worth a closer look.

That distinction matters for any practice. Forwarding a retired address passes visitors and search signals to the page that replaced it. Asking Google to index a thin or duplicate page rarely helps, and fixing the page itself usually comes first.

How every change was checked

Each change was drafted, reviewed against the practice’s confirmed facts, and approved before it went live. Clinical details, such as which ages are seen, which conditions are treated and who provides care, were checked against what the practice actually offers, not against what would read well.

After publication we checked the live pages rather than the drafts: titles in the page source, links that resolved, and forwarded addresses that answered with a permanent redirect to the right page. Testing the live result is the step most often skipped, and it is where small mistakes are caught.

What we will watch, and when

Search engines usually need several weeks to re-crawl changed pages and settle their positions, so the first comparison will use a full period after the changes against a matching period before them. Looking sooner tends to show noise, not a trend.

The comparison will track the searches each change was aimed at: did the Knoxville depression page start appearing for “depression treatment Knoxville”, did click-through rise on the retitled pages, and did the forwarded addresses drop out of the not-found list. Whatever the numbers show, they will be added to this page with their dates.

How to repeat this on your own practice’s site

Export the last three months of your Search Console performance data and filter for searches with many impressions and an average position between eight and twenty. List the page each one lands on. For each, compare the page title and opening paragraph with the words in the search, and note what the page leaves unanswered.

Then open the indexing report and separate retired addresses from live pages. Forward the retired ones to their closest replacement, and improve the live ones before requesting indexing. Make one change at a time where you can, write down the date, and give each change weeks, not days, before judging it.

How results will be measured

The changes went live in late October 2026, so there are no results to report yet, and we will not guess. Search engines take weeks to re-crawl and re-rank pages.

We will compare the same Search Console measures, clicks, impressions, click-through rate and average position, for the same searches, over matching periods. We will publish the numbers here whether they improve or not.