The Hillcane Title and Meta Rewrite Program: Results
An owner asked me recently whether rewriting titles and meta descriptions across a whole site is actually worth the work, or whether it just feels productive. Short answer: it is worth it for consistency and cleanliness, and it is too early to tell you whether it changed clicks, because our own site is only days old and most posts have no meaningful before period to compare against.
This post reports the Hillcane rewrite program plainly, as a method and a set of counts, not as a traffic win. We ran nine title batches on September 19, 2026 that came to roughly 858 posts total, followed by a meta description pass of 445 written on September 20, 2026, with 49 retries cleared after a batch of oversized schema tags needed to be split apart. After the last batch, a set of guard checks confirmed the state of the titles across the site.
I am writing this the way I would want to read it if I were the owner asking the question: what the rules were, what actually happened batch by batch, what the checks found, and an honest statement of what we still do not know because the window since launch is too short to read.
The Rules
Before any batch ran, we set six rules for every rewritten title and meta description, and held every batch to the same six rules rather than improvising case by case. Titles needed to fall within a defined length window rather than running long or unusually short, no hyphens were allowed anywhere in a rewritten title, city and state were added where they fit naturally rather than forced into every single post, and slugs were left untouched throughout the entire program so that no published URL would change underneath an existing link.
The remaining two rules covered consistency between what a post's own heading said and what its search facing title tag said, and a check that no two titles across the site ended up identical to each other. Holding every batch to the same six rules, rather than tightening or loosening them as the program went, is part of what made the guard checks after the final batch meaningful rather than a moving target.
None of these six rules is exotic. They are close to what Google Search Central's own guidance on influencing title links in search results describes as sensible practice: clear, descriptive, reasonably concise titles that avoid unnecessary length and avoid duplicating text across pages. We applied a plain, consistent set of rules across a real, large batch of posts and then checked our own work.
The Batches
The program ran as nine separate batches, all applied on September 19, 2026, sized at 24, 100, 90, 87, 160, 95, 100, 100 and 102 titles respectively. Summed together, those nine batches total 858 titles applied, which is the figure we cite as the program's overall size. Running the same nine numbers back through addition confirms the total: 24 plus 100 plus 90 plus 87 plus 160 plus 95 plus 100 plus 100 plus 102 comes to 858.
Our own tracker holds itemized detail, meaning the specific old title and new title, for 465 of the posts touched across the program, and of those, 218 have both an old title and a new title recorded, of which 215 changed and 3 were confirmed as already meeting every rule and left as is. Batches four and five, and batches six through nine, do not carry full old and new title pairs in the tracker file the way the earlier batches do, so the itemized comparison below is drawn specifically from that smaller, fully paired set of 215 changed titles rather than from all 858.
Within that set of 215 changed titles, the median length of the old title was 94 characters, and the median length of the new title was 57 characters, a median drop of 38 characters per title. None of the new titles in that set ran over 62 characters. Looking at the shape of the rewritten titles themselves, the large majority, 200 of them, followed a topic followed by a colon followed by a short payoff phrase pattern, with 11 written as plain descriptive titles and 4 written as a direct question.
The Meta Description Pass
A separate pass covered meta descriptions rather than titles, run on September 20, 2026, one day after the title batches finished. It wrote 445 meta descriptions across the site. During the run, 49 posts initially failed and needed a retry, and the cause was identified plainly: those posts had oversized schema tags that needed to be split before the meta description write could complete correctly. After splitting the oversized schema, all 49 of those retries completed successfully, for a final result of 49 out of 49 resolved.
That kind of failure and fix, rather than a silent skip, is worth stating clearly, because a data study is only as trustworthy as its honesty about what did not work on the first attempt. The oversized schema issue was specific and identifiable, not a mystery, and once it was found the fix was straightforward and the retries cleared completely.
It is worth pointing out that the meta description pass ran as its own separate step, one full day after the title batches finished rather than in the same pass. Keeping titles and meta descriptions as two distinct steps, rather than one combined rewrite, made it easier to isolate the schema related failure to the meta description pass specifically, rather than having to untangle it from any title related issue at the same time. A single combined pass covering both fields at once might have made that same failure harder to trace back to its actual cause.
The Guard Checks
After the last of the nine title batches, we ran a set of guard checks across the site to confirm the rules had held, rather than assuming they had because the rules existed on paper. The checks confirmed no titles over 70 characters remained anywhere on the site, no titles under 30 characters remained, no two titles were duplicated across different posts, and there were zero mismatches between a post's own heading and its search facing title tag.
Those four checks map directly back to the six rules set before the program began. A clean result on all four, checked after the final batch rather than after each individual batch, is what lets us say the program held its own standard across the full run rather than drifting as it went.
What This Program Is Actually Worth
The honest position here is that the value of a program like this is consistency and cleanliness first, not a proven traffic result. We rewrote titles that were, in the paired sample, a median of 94 characters down to a median of 57, well within a length that Google is far less likely to truncate or rewrite on its own. We closed the gap between what a post says at the top of the page and what its title tag says in search results down to zero mismatches. Those are real, measured outcomes of the program itself.
What we cannot yet say is whether any of this changed impressions or clicks, because hillcane.co launched September 9, 2026 and the title batches ran on September 19, 2026, which leaves too short a window, and too little of a clean before period for most posts, to read any performance change honestly. Saying otherwise would be exactly the kind of invented confidence this whole project tries to avoid.
If you are an owner reading this wondering whether a similar rewrite pass is worth doing on your own site, the honest answer is that the process itself, clear rules held consistently, guard checks run at the end, and an honest accounting of what failed and got fixed, is worth doing on its own terms. A cleaner, more consistent set of titles and descriptions is a reasonable goal by itself, separate from any claim about what it will do to your traffic.
It is also worth saying that a site with a longer history than ours would be a better place to measure a change like this cleanly, since it would have a real before period to compare against rather than a handful of days. That is simply not the situation hillcane.co is in yet, and pretending otherwise for the sake of a tidier story would not serve any owner reading this looking for an honest account of what a rewrite program can and cannot tell you this early.
A Repeatable Routine, Not a One Time Project
What this program leaves behind, beyond the counts themselves, is a routine any site could run again later: fix rules first, batch the work into sized groups rather than attempting everything at once, keep a tracker with old and new values wherever practical, and finish with guard checks that verify the rules held rather than trusting that they did.
The batch sizing itself is worth noting as a practical detail. Running 858 titles as nine batches of varying size, from 24 up to 160, rather than as one single pass, made it possible to check work partway through and catch any drift in the rules before it spread across the whole site. That structure is repeatable for any future pass, whether that is a second round of titles months from now or a first round for a newer, smaller site.
We will revisit performance once there is a real window to read, and when we do, that will be reported the same way this post has tried to: with the dates, the sample size, and an honest line about what the data does and does not show.
Why the Batch Sizes Varied
Looking at the nine batch sizes side by side, 24, 100, 90, 87, 160, 95, 100, 100 and 102, the shape is worth noticing even without a stated reason for each individual number. The first batch was the smallest, at 24, and the rest mostly ran close to 90 to 100 posts each, with one larger batch of 160 in the middle of the sequence. That shape, a small first pass followed by steadier, similarly sized groups, is a reasonable way to run a large rewrite without committing the full site to an unproven set of rules in one single pass, though the exact reasoning behind each specific batch size is not recorded in the tracker itself and is not something this post will guess at beyond what the numbers show.
What matters more than the reasoning behind any single batch size is that every batch, regardless of how large, was held to the same six rules and checked against the same guard checks at the end. A program that varies its batch sizes but not its standards is still one consistent program, and that consistency across nine separate runs on a single day is part of what the final guard check results are confirming.
None of this changes the total. Nine batches, run on one day, held to one set of rules, summing to 858 titles, is the shape of the program regardless of why any individual batch landed at the size it did.
Related Reading
More from Hillcane on this topic, plus the pages on the site that sit next to the work.
Frequently Asked Questions
What did the rewrite program change?
Nine title batches applied on September 19, 2026 covering roughly 858 posts, followed by a meta description pass of 445 written on September 20, 2026 with 49 retries cleared after oversized schema tags were split. Guard checks after the last batch found no titles over 70 characters, none under 30, no duplicates, and zero mismatches between post headings and title tags.
How many titles were rewritten?
The nine batches applied on September 19, 2026 total 858 titles, sized at 24, 100, 90, 87, 160, 95, 100, 100 and 102. Our tracker holds itemized detail for 465 of those posts, and 218 of them carry both an old and a new title. Of that paired group, 215 changed and 3 were confirmed as already meeting every rule and left alone. The remaining 247 itemized posts lack full old and new pairs.
What did the meta description pass find?
It wrote 445 meta descriptions on September 20, 2026, one full day after the title batches finished. Along the way, 49 posts failed on the first attempt because oversized schema tags had to be split before the write could complete. Once those tags were split, all 49 retries cleared, a final result of 49 out of 49. Running descriptions as a separate step made that failure easy to trace to its actual cause.
Did this program improve rankings or clicks?
No such claim is made here. Hillcane launched September 9, 2026, and the title batches ran on September 19, 2026, so the window is too short and most posts have no clean before period to compare against. What the program can show is measured process: 858 titles applied, 445 meta descriptions written, and guard checks passed. Performance will be revisited and reported with dates and sample sizes once a real window exists.
What length did titles typically drop to?
Titles got much shorter. In the fully paired sample of 215 changed titles, the median old title ran 94 characters and the median new title ran 57 characters, a median drop of 38 characters. Not one rewritten title in that set exceeded 62 characters. That puts new titles well inside a length Google is far less likely to truncate or rewrite on its own, which fits Google Search Central guidance on concise, descriptive title links.
What were the six rules the program followed?
Six rules applied to every batch. Titles had to fall within a defined length window, and no hyphens were allowed anywhere in a rewritten title. City and state went in where they fit naturally rather than being forced. Slugs stayed untouched so no published URL changed. A post's heading and its title tag could not mismatch, and no two titles on the site could be identical. None of the six changed between batches.
What shape did most of the rewritten titles take?
Most followed a topic, then a colon, then a short payoff phrase. Among the 215 changed titles with full old and new detail in the tracker, 200 used that pattern, 11 were written as plain descriptive titles, and 4 were phrased as a direct question. Each of those shapes still had to satisfy the same six rules, including the length window, the ban on hyphens, and no duplicates across the site.
What caused the meta description retries?
Oversized schema tags caused them. During the September 20, 2026 meta description pass, 49 posts failed on the first try because the schema attached to them was too large and had to be split apart before the description write could finish. After the split, all 49 resolved on retry. Because descriptions ran one day after the title batches, the problem could be isolated to the description run instead of being tangled up with titles.
What did the guard checks confirm after the last batch?
Four checks came back clean after the ninth and final batch. No title anywhere on the site ran over 70 characters, none fell under 30 characters, no two posts shared a duplicate title, and there were zero mismatches between a post's own heading and its search facing title tag. Those checks map back to the six rules, and running them once at the end confirms the standard held across all 858 titles.
How can Hillcane help with a similar rewrite program on another site?
Hillcane can help you set clear rules first, split the work into sized, checkable batches, keep a tracker of old and new values, and run guard checks at the end to confirm the rules held. That is the same routine behind 858 titles and 445 meta descriptions here. To talk through a similar program for your own site, reach out at hillcane.co/contact or call (256) 384-2449.
Work with Hillcane
Nine batches, roughly 858 titles rewritten, a meta pass of 445, and guard checks that came back clean. Here is exactly what the Hillcane rewrite program did, and what it is still too early to claim.
If you want a similar rewrite pass run on your own site, with the same rules and the same guard checks, reach Hillcane at hillcane.co/contact or call (256) 384-2449.
Comments