Amrut Services

Mehul Dedhia

SEO Consultant · Amrut Services

Category: SEO

  • Do manufacturer descriptions need rewriting?

    Do manufacturer descriptions need rewriting?

    TL;DR

    Usually not, and the standard advice to rewrite all of them is why so many catalogues never get anything done. There is no duplicate content penalty. You just cannot win a comparison you never entered. Rewrite ten to thirty, not three thousand, and never paraphrase what the manufacturer already wrote.

    Usually not, and the standard advice to rewrite all of them is the reason so many catalogues never get anything done.

    Here is the actual answer: rewrite the ones that can rank, leave the rest, and delete the ones that should not exist. Which is a different job from rewriting three thousand product pages, and a job somebody can start on Monday.

    What the duplication actually costs

    Less than people think, and not what they think.

    There is no penalty for using a supplied description. Google does not punish it. What happens is quieter: your page is one of several hundred containing the identical text, so it has nothing to distinguish it, and Google picks whichever of those pages it considers most trustworthy. That is usually the manufacturer or the largest retailer, and it is not you.

    So the cost is not a penalty. It is that you cannot win a comparison you have not entered.

    And it is a crowded field by construction. HTTP Archive’s 2025 Web Almanac found ecommerce software on 19.2% of mobile sites. A very large number of stores are listing the same items with the same supplied text, and being the four-hundredth copy is not a position improved by paraphrasing.

    Where the “add 500 words” advice goes wrong

    I have written a version of it myself in older audits and I would write it differently now.

    It is not wrong so much as unactionable, and unactionable advice does real damage. The owner hears three thousand descriptions, concludes it is impossible, and does nothing at all. So the twenty pages that could genuinely have ranked stay untouched, and the template fix that would have helped all three thousand never happens either.

    Note also what Google’s spam policies say about content produced at scale: the definition of scaled content abuse turns on pages “generated for the primary purpose of manipulating search rankings and not helping users”, and contains no word count anywhere. Three thousand paraphrased descriptions written to hit a length are closer to that definition than to a solution.

    Which ones to rewrite

    The same filter I would apply to any product page. Does anybody search for this specific item. Can you say something the manufacturer cannot. Will it still be on sale in six months. Would you back it with internal links.

    On most catalogues that is ten to thirty products, not three thousand. Thin at scale is rarely a writing problem. It is a decision nobody made about which pages deserve the effort.

    What to write, when you do

    Not a paraphrase. Paraphrasing is the worst possible use of the time. It costs as much as writing and adds nothing a customer wants.

    What a retailer can say that a manufacturer cannot:

    • Fit and comparison. How it compares to the other three you stock. Which one you would recommend and for whom. The manufacturer cannot write that sentence.
    • What you have learned from selling it. What gets returned and why. What people ask before buying. Which size runs small.
    • Practical specifics. Whether it fits the thing customers actually want it for. Photographs of the real item rather than the supplied render.

    Two honest paragraphs of that beat eight hundred words of restatement, and it is content nobody can copy because it comes from handling the stock.

    What the other 2,970 need

    Keep the supplied description. It is accurate and customers do read it.

    Add the template-level baseline that costs nothing per product: a unique title, a real H1, correct structured data, alt text, and a breadcrumb linking up to a category that is a ranking project. One change, whole catalogue.

    Then accept that those pages exist to convert people who arrive, not to attract them. That is a legitimate job for a page.

    The short version

    1. There is no duplicate content penalty. You just cannot win a comparison you have not entered.
    2. 19.2% of mobile sites are ecommerce, and the same text is on hundreds of stores.
    3. “Rewrite everything” is unactionable, so nothing gets done at all.
    4. Google’s scaled content definition has no word count in it.
    5. Rewrite ten to thirty, not three thousand.
    6. Never paraphrase. Write fit, comparison and what you learned from selling it.
    7. Everything else gets the template baseline and converts rather than attracts.

    If a site should be ranking and it isn’t, that’s the work I do. Ecommerce SEO covers catalogues, categories and what to write where. If you’re not sure which of several plausible problems is costing you, that’s what an SEO audit is for.

  • What to measure when rankings are not the goal

    What to measure when rankings are not the goal

    TL;DR

    A client once told me they were first for their main term and the phone was not ringing. Both were true and only the second mattered. Rankings were a reasonable proxy when a result page was ten blue links. Sixty-eight per cent of searches now end without a click. Report enquiries first.

    A client once told me they were first for their main term and the phone was not ringing. Both statements were true, and the second one is the only one that mattered.

    Rankings are a proxy. They were a reasonable proxy when a result page was ten blue links and position one meant most of the clicks. They are a much weaker proxy now, and reporting them as the outcome quietly lets everyone avoid the question of whether the work produced anything.

    Why the proxy broke

    Three things happened, and they compound.

    The result page stopped sending clicks. SparkToro’s analysis of Similarweb’s US clickstream found 68.01% of Google searches ended without a click in early 2026, up from 60.45% in 2024, with AI Overviews on over 20% of searches cutting click-through by roughly 60% where they appear. Position one on a query that sends nobody anywhere is a screenshot, not a result.

    Position stopped meaning one thing. Google’s own documentation defines average position as the topmost position you occupied, averaged across the queries where you appeared, and warns it is “a complex metric that can be misleading if you don’t understand the subtleties.” Position three under a map pack, an AI overview and four ads is not position three as anyone imagines it.

    Personalisation and location. There is no single ranking to report. The number in the tool is one sample.

    What I would report instead

    In rough order of how close each sits to the money.

    • Enquiries, by source. Calls, forms, bookings. If you cannot attribute them, count them anyway. A total that moves is more informative than a position that does not.
    • Clicks, not impressions. Impressions measure how often you appeared, which is partly a function of how deep result sets ran. Twenty-five thousand impressions on a query with no commercial relevance is not an opportunity, and it will make a chart look excellent.
    • Share of clicks on the queries you care about. Pick the twenty or thirty terms that describe your actual business and track clicks on those as a set. This is the closest honest substitute for “are we winning” and it ignores the thousands of irrelevant terms inflating your totals.
    • Pages entering the index that were meant to. A blunt but useful health metric on any site publishing regularly.
    • Speed to first response. Not an SEO metric and it belongs on the same page as them, because it decides what the traffic converts to. MIT’s study of 15,000 leads and 100,000 call attempts found the odds of qualifying a lead falling twenty-one-fold between a five-minute and a thirty-minute response.

    What I would stop reporting

    Keyword count. “Ranking for 26,618 keywords” sounds like progress. I have seen a site ranking for ten thousand more terms than its competitor and converting almost none of them into positions anybody sees. Hit rate beats presence every time.

    Domain authority scores. Third-party inventions, useful for rough comparison and meaningless as a target. Reporting them monthly trains the client to care about a number no search engine uses.

    Average position across everything. An average over a mixed bag of relevant and irrelevant queries, where losing bad rankings improves the number.

    Anything the client cannot act on. If a number moving would not change what anybody does next week, it is decoration.

    The awkward part

    These metrics are less flattering. Positions improve steadily and enquiries do not. They jump, stall, and depend on things I do not control, like whether anybody answers the phone.

    Reporting them means occasionally showing a month where the work went well and the number did not move. I think that is the right trade, and it is the reason most reporting avoids it.

    The short version

    1. Rankings are a proxy and the proxy weakened. 68% of searches end without a click.
    2. Position one on a zero-click query is a screenshot.
    3. Google warns average position misleads. It is topmost-position-averaged.
    4. Report enquiries first, even unattributed.
    5. Clicks on a chosen query set beats totals inflated by irrelevant terms.
    6. Put speed-to-response on the same page. 21-fold difference between five and thirty minutes.
    7. Drop keyword counts and authority scores. If nobody acts on it, it is decoration.

    If a site should be ranking and it isn’t, that’s the work I do. SEO consulting covers what to measure and what to ignore. If you’re not sure which of several plausible problems is costing you, that’s what an SEO audit is for.

  • Seasonal dip, or a real problem?

    Seasonal dip, or a real problem?

    TL;DR

    Traffic is down eighteen per cent on last month. That number is worth nothing on its own, and it is the number most reports lead with. Month over month is a random number generator that produces a panic about four times a year. Compare the same month last year, over two years if you have them.

    Traffic is down eighteen per cent on last month. That number, on its own, is worth nothing, and it is the number most reports lead with.

    Almost every business has a shape to its year. Until you know yours, month-over-month comparison is a random number generator that produces a panic roughly four times a year.

    The comparison that actually works

    Same month, previous year. Not last month. Not the rolling average.

    If November is always down on October, November being down on October tells you nothing. November being down on last November tells you something, and it is the only comparison that controls for the shape of your year.

    Two years of overlay is better than one, because a single previous year could itself have been unusual.

    The trap in that, though

    Year-over-year has become less reliable than it used to be, and it is worth knowing why before you lean on it entirely.

    The share of searches ending without a click has been rising steadily. SparkToro’s analysis of Similarweb’s US clickstream put it at 68.01% in early 2026, against 60.45% in 2024.

    So a site holding all of its positions can be down year over year on clicks, every year, for reasons that have nothing to do with the site. A modest annual decline in clicks against flat impressions is now close to the baseline condition, and treating it as a fault sends people looking for a problem that is not there.

    Which is why I check impressions alongside clicks before concluding anything. Those two columns answer different questions and the difference between them is usually the actual finding.

    Things that look seasonal and are not

    A competitor who launched in your quiet season. You notice in the busy one, and it reads as a bad year rather than a new entrant.

    A slow decline with a seasonal shape on top. The seasonality masks the trend. Plot twenty-four months rather than twelve and the underlying line becomes visible.

    Something that broke during the quiet period. Nobody noticed because traffic was low anyway, and it surfaces as a disappointing peak.

    Things that are seasonal and get treated as problems

    School holidays for anything sold to parents or professionals.

    The genuine off-season in trades, tourism, and anything weather-driven.

    Christmas and the fortnight after it, which distorts a December-to-January comparison in nearly every B2B account.

    The reporting month itself. A month with fewer working days, or five weekends rather than four, moves B2B numbers by several per cent for no reason at all.

    And be careful about which pages you are averaging across. Ahrefs looked at 14 billion pages and found 96.55% get zero traffic from Google. On most sites a handful of pages carry nearly everything, so a “site-wide dip” is frequently one page moving. Segment before you diagnose, or you will investigate a whole site to find one URL.

    How I would set this up once

    • Build the seasonal baseline before you need it. Twenty-four months of clicks and impressions, plotted. Half an hour, once. Every future conversation about a dip refers to it.
    • Annotate it. Launches, migrations, campaigns, confirmed Google updates. A dated line on a chart resolves most arguments about causation in seconds.
    • Report year over year by default, with month over month as secondary. This one change removes most false alarms.
    • Set a threshold in advance. Decide what size of year-over-year move is worth investigating before you are in the middle of one. Otherwise every number becomes a discussion.

    The honest limit

    Sometimes there is not enough history. A site eighteen months old has no reliable baseline and you have to say so rather than manufacture confidence. In that situation I would watch impressions, watch the competitor set, and be explicit that the first full year is the one that produces the baseline rather than the answers.

    The short version

    1. Month over month is noise. Same month, previous year.
    2. Two years of overlay beats one.
    3. Zero-click has risen from 60% to 68% in two years, so a modest annual click decline can be the baseline, not a fault.
    4. Check impressions alongside clicks before concluding.
    5. Plot twenty-four months to see a trend hiding under seasonality.
    6. Annotate launches, migrations and confirmed updates.
    7. Set the investigate-threshold in advance, not mid-panic.

    If a site should be ranking and it isn’t, that’s the work I do. SEO recovery covers telling a real decline from a normal one. If you’re not sure which of several plausible problems is costing you, that’s what an SEO audit is for.

  • Traffic dropped overnight: the first six things I check

    Traffic dropped overnight: the first six things I check

    TL;DR

    The call always has the same shape. Traffic fell off a cliff, nobody changed anything, and it must be Google. In fifteen years it has almost never been Google. Check whether the drop is real before anything else, then ask what shipped that week, and ask the developer rather than the marketer.

    The call always has the same shape. Traffic fell off a cliff, nobody changed anything, and it must be Google.

    In fifteen years it has almost never been Google. Here is the order I work through, which is roughly the order of how likely and how cheap each explanation is.

    1. Is the drop real, or is it the tracking?

    First, always, and it resolves a meaningful share of these calls in ten minutes.

    A consent banner change, a tag manager container republished, a plugin update, a migration that dropped the analytics snippet from one template. Any of those produces a vertical cliff in analytics with nothing whatsoever happening in search.

    The test: does Search Console agree? If clicks in Search Console are flat while analytics has collapsed, you have a measurement problem, not a traffic problem. If both fell, keep going.

    2. What shipped that week?

    Before touching anything, get the exact date from Search Console and ask what happened in the days before it. A release, a theme update, a plugin, a migration, a robots.txt edit, a new SEO plugin with opinionated defaults.

    “Nobody changed anything” is said in good faith and is wrong more often than it is right, because the person saying it did not personally change anything. Ask the developer, not the marketer.

    3. Is something blocking or noindexing at scale?

    The catastrophic causes are few and quick to rule out. A staging Disallow: / that shipped. A noindex applied at category level. A canonical pointing somewhere wrong across a template.

    Check the raw source and the HTTP headers, because X-Robots-Tag does the same job invisibly. This is the same sequence I run for pages that dropped out of the index, and on a sitewide drop it is the first thing that would explain the shape.

    4. Did impressions fall, or only clicks?

    This is the diagnostic split people skip, and it separates two completely different problems.

    Impressions fell. You lost visibility. Rankings, indexing, or a competitor.

    Impressions held and clicks fell. You are still appearing and fewer people are clicking. That is a result-page change, not a ranking change, and it is increasingly the answer. SparkToro’s analysis of Similarweb’s US clickstream found AI Overviews now appearing on over 20% of searches and cutting click-through by roughly 60% where they appear.

    A site can hold every position it had and lose a third of its clicks. No amount of technical work fixes that, and diagnosing it as a ranking problem wastes a quarter.

    It is the same split that separates the four causes of a high-impression, low-click row. There the question is why one query underperforms; here it is why the whole site did. Same two columns, same discipline: read them apart before you read them together.

    5. Read the position column properly

    Average position is not what most people think it is. Google’s own documentation defines it as the topmost position your property occupied, averaged across queries where you appeared, and states plainly that it is “a complex metric that can be misleading if you don’t understand the subtleties.”

    Two consequences that matter during a drop. Losing rankings on weak queries can make average position improve, because the bad ones stopped appearing. And a position that appears stable can hide a large change in which queries produced it.

    Compare query sets, not averages.

    6. Only now, consider Google

    A core update, or an algorithmic change. Real, and last on the list because it is the explanation people reach for first and it is rarely correct.

    The tells: it started on a date Google has confirmed, it affected a category of pages rather than the whole site, and competitors moved too. If your drop is sitewide, overnight, and your competitors are unchanged, that pattern does not look like an update. It looks like something you shipped.

    What I would not do in week one

    • Rewrite content. If the cause is a header directive, a month of writing changes nothing and you will conclude content does not matter.
    • Mass-resubmit for indexing. It treats a symptom and tells you nothing.
    • Panic-disavow. Almost never the answer and occasionally makes things worse.

    The short version

    1. Check the tracking first. Does Search Console agree with analytics?
    2. Ask what shipped that week, and ask the developer, not the marketer.
    3. Rule out sitewide noindex, robots block and canonical faults. Check headers too.
    4. Impressions or clicks? Different problems entirely.
    5. Clicks alone can fall from AI Overviews, on 20%+ of searches, cutting CTR ~60%.
    6. Average position is topmost-position-averaged and Google warns it misleads.
    7. Consider a core update last. Sitewide and overnight usually means you shipped something.

    If a site should be ranking and it isn’t, that’s the work I do. SEO recovery covers diagnosing drops and what realistically comes back. If you’re not sure which of several plausible problems is costing you, that’s what an SEO audit is for.

  • Reading Business Profile insights when the catchment is three towns

    Reading Business Profile insights when the catchment is three towns

    TL;DR

    A clinic sitting where three towns meet has a reporting problem before it has a marketing one. One profile reports one set of numbers for a catchment that behaves like three markets. The Search and Maps split is not pack versus organic, and a grid check beats the report for anything spatial.

    A clinic sitting where three towns meet has a reporting problem before it has a marketing one. The profile reports one set of numbers for a catchment that behaves like three different markets, and most of the conclusions people draw from that report are wrong.

    Here are the traps, in the order I hit them.

    Trap one: the Search versus Maps split is not what you think

    This is the big one and I have watched a client move budget on the strength of misreading it.

    Your profile insights split how people found you between Google Search and Google Maps. Almost everybody reads that as organic versus map pack, sees Search at seventy or eighty per cent, and concludes the pack is not working.

    That is not what it measures. It is where the profile was viewed: the Maps app and maps.google.com on one side, google.com on the other. The local pack renders inside Google Search. So a profile getting most of its views through Search is very often getting them from the pack, sitting at the top of the results page.

    A high Search share is not evidence of a pack problem. As a pack-versus-organic proxy the ratio is close to meaningless, and it is the most confidently misread number in the whole report.

    Trap two: one profile cannot tell you about three towns

    The insights aggregate. They will tell you calls went up and give you no way to know which of your three towns produced them.

    Which matters because distance is one of Google’s three local ranking factors and it is measured from your building. You are almost certainly strong in the town you sit in, competitive in the next, and invisible in the third, and the aggregate number hides all three of those facts behind one line.

    The fix is not in the report. It is a grid check: run the same search from several points across the catchment and see where you actually appear. Twenty minutes, and it tells you what the insights structurally cannot.

    Worth knowing what you are measuring against, too. Google names relevance, distance and prominence as the three local ranking factors and states plainly that there is “no way to request or pay for a better local ranking on Google.” Two of those three you can work on. The third explains most of what the grid will show you.

    Trap three: click mix varies enormously and the benchmark is useless

    Across six local profiles I track, the share of profile clicks going to the phone ranged from 13.1% at one business to 3.1% at another. Same report, same metric, same country, wildly different meaning.

    Direction requests are close to meaningless for some businesses and the primary signal for others. So a single cross-client scorecard misleads, and a percentage you read in an industry benchmark tells you nothing about whether your own mix is healthy.

    Compare your profile to its own past. That is the only comparison available that means anything.

    What I would actually track

    • Calls, direction requests and website clicks, month over month, against your own baseline. Absolute numbers, not percentages.
    • Seasonality before conclusions. A clinic has term-time and holiday patterns. A dip in August is not a ranking event.
    • Grid position, quarterly. The thing the insights cannot give you.
    • Review recency alongside it. BrightLocal’s 2026 survey of 1,002 US consumers found 74% look for reviews written in the last three months, so a flat month with a stale profile is a different diagnosis from a flat month with a fresh one.

    The honest limit

    Profile insights are a description of what happened, not an explanation. They will tell you calls fell. They will never tell you whether that was a competitor opening nearer, a category change, a seasonal pattern, or nothing at all.

    I would treat them as the thing that prompts a question rather than the thing that answers it, and I would not report them to a client as though a number moving is a finding.

    The same caution applies to the site side of the ledger. If the website never answers the question every patient has before booking, no amount of profile optimisation fixes the drop-off, and the insights will show you a healthy view count right up until the moment somebody closes the tab.

    The short version

    1. The Search/Maps split is where the profile was viewed, not pack versus organic. The pack lives inside Search.
    2. One profile cannot report on three towns. The aggregate hides the geography.
    3. Grid check beats the report for anything spatial. Twenty minutes.
    4. Click mix varies hugely: 13.1% phone share at one business, 3.1% at another. Benchmarks are useless.
    5. Compare the profile to its own past.
    6. Check seasonality before calling a dip a problem.
    7. Insights prompt questions. They do not answer them.

    If a site should be ranking and it isn’t, that’s the work I do. Chiropractic SEO covers the profile, the catchment and what the numbers actually mean. If you’re not sure which of several plausible problems is costing you, that’s what an SEO audit is for.

  • Three videos a practice actually needs

    Three videos a practice actually needs

    TL;DR

    Video advice for clinics is usually a content calendar, which is why practices try it for a month and stop. Three videos, made once, do most of the available work. The first appointment walkthrough is the highest-value one and almost nobody has it. Buy the microphone, not the camera.

    Video advice for clinics is usually a content calendar: post weekly, build a channel, be consistent. Most practices try it for a month and stop, which is the correct response to advice that costs more than it returns.

    Three videos, made once, do most of the available work. Here they are.

    They also close one of the gaps that turns up in every local audit I run. Writing about what clinic websites leave out, the pattern is consistent: the money question, the reviews page, and video. Three things, none expensive, none present.

    First, why it is worth anything at all

    Not because Google owns YouTube. That reasoning gets repeated constantly and it does not hold. Ownership is not why video surfaces.

    The real reason is unoccupied ground. HTTP Archive’s 2025 Web Almanac found a video element on just 5.7% of mobile pages. Ninety-four per cent of the web has none. For a clinic competing against four others in the same suburb, that is not a box to tick. It is the one differentiator on the list that nobody local is likely to have taken.

    The second reason is that for a treatment nobody has had before, ninety seconds of the practitioner explaining what happens does more for a booking decision than a page of text. It is the cheapest trust signal available to a small clinic: a real person, in the real room, saying the thing.

    Trust is doing most of the work in this decision. BrightLocal’s 2026 Local Consumer Review Survey found 97% of consumers read reviews for local businesses and 31% will only use one rated 4.5 or above. Reviews carry the bulk of that, and video carries the part reviews cannot: what the room looks like and whether the person seems like someone you would let treat you.

    The three

    1. What happens at your first appointment. Filmed in the actual clinic, by the actual practitioner, ninety seconds. Where to park, what reception looks like, what you will be asked, whether you undress, what the first adjustment feels like, how long you will be there.

    This is the highest-value video a clinic can make and almost none of them have it. New patient anxiety is the main thing standing between an interested person and a booking, and it is almost entirely about the unknown physical experience. This removes it.

    2. The practitioner explaining the condition you treat most. Not a lecture. The same explanation you give in the room, ten times a week. What is happening, why it hurts, what treatment does about it.

    Put it on the matching condition page, where somebody is already reading about that exact problem. The video does not need to rank on its own to earn its place there.

    3. A patient talking, if you can get one. Harder, and worth the effort because it is the one competitors cannot copy. Not scripted, not polished. Someone saying what they came in with and what happened.

    If patient privacy or reluctance makes this impossible, replace it with the practitioner answering the money question directly on camera. Less powerful, and still better than the page that does not mention cost.

    What matters and what does not

    Does not matter: production quality beyond decent light and audible sound, editing, music, a channel banner, posting schedule, subscriber count.

    Does matter: that it is genuinely the practitioner, that it is filmed in the room the patient will walk into, and that it is embedded on the page where the relevant decision is being made.

    A phone on a tripod near a window, with a cheap lapel microphone, is sufficient. The lapel microphone is the one thing worth buying, because people forgive poor picture and abandon poor sound.

    Where to put them

    On the site first, on the relevant page. The first appointment video belongs on the new patient and contact pages. The condition explanation belongs on the condition page.

    Upload to YouTube as well, because it is free surface area and it feeds the Business Profile, but the primary job is happening on your own pages where the person is already deciding.

    And embed something on the profile. In three clinic audits I ran in one year, none of the three had a YouTube channel and none had video anywhere on the site. Three out of three.

    The short version

    1. Not because Google owns YouTube. That reasoning is wrong.
    2. 5.7% of mobile pages have a video element. It is unoccupied ground locally.
    3. Three videos, made once. Not a content calendar.
    4. First appointment walkthrough is the highest-value one and almost nobody has it.
    5. The condition explanation goes on the condition page.
    6. A patient talking, if you can get one. Otherwise the money question on camera.
    7. Buy the microphone, not the camera.

    If a site should be ranking and it isn’t, that’s the work I do. Chiropractic SEO covers the site, the profile and the trust signals around them. If you’re not sure which of several plausible problems is costing you, that’s what an SEO audit is for.

  • A treatment page per condition, not a ‘Services’ page

    A treatment page per condition, not a ‘Services’ page

    TL;DR

    Nobody wakes up wanting chiropractic. They wake up with sciatica, or headaches three times a week, and they search that. Most clinic sites list adjustments, cupping and dry needling, which describes what you do rather than what is wrong with them. Six to ten condition pages, not thirty.

    Nobody wakes up wanting chiropractic. They wake up with sciatica, or a neck that will not turn, or headaches three times a week, and they search that.

    Most clinic sites have a Services page listing adjustments, soft tissue work, cupping, dry needling and rehab. That page describes what you do. The person searching is describing what is wrong with them, and those are different vocabularies.

    Techniques versus conditions

    This is the whole distinction and it is worth stating plainly.

    Technique pages, meaning spinal manipulation, myofascial release and cupping, are written from inside the practice. They assume the reader already knows they want that thing.

    Condition pages, meaning sciatica, whiplash, tension headaches, lower back pain in pregnancy and shoulder impingement, are written from the patient’s side. They meet the search that is actually being typed.

    You need a small number of the first for people comparing modalities. You need the second for everybody else, and most clinics have it backwards: five technique pages and no condition pages at all.

    What a condition page has to do

    Answer the questions in the order the person has them, which is almost never the order a clinician would choose.

    • What is probably causing this. In plain language. Not a paragraph of anatomy.
    • Is it serious. The unspoken question underneath every symptom search, and the one nobody answers. Saying clearly when someone should see a doctor rather than a chiropractor is the most trust-building sentence you can write, and it costs you almost no business. The people it turns away were not your patients.
    • What treatment actually involves. What happens in the first appointment, physically. People are nervous about being adjusted and nobody tells them what it feels like.
    • How long it takes. Number of visits, honestly. This is the question every patient has and almost no clinic site answers.
    • What it costs and whether insurance covers it. The question every patient has before they book, and the one most clinic websites leave out entirely.

    That last one is not a minor omission. BrightLocal’s 2026 survey of 1,002 US consumers found 97% read reviews for local businesses before choosing. People research healthcare decisions thoroughly, and a clinic that answers the money question on the page is removing the reason to go and research somebody else instead.

    Which conditions

    The ones you actually treat and see regularly. For most clinics that is six to ten: lower back pain, sciatica, neck pain, headaches, whiplash, shoulder pain, and whatever the practice specialises in: pregnancy, sports injury, paediatric.

    Not thirty. A page for every condition in the textbook is a set of thin pages competing with each other, and the honest test is whether you could write it from your own experience of treating it rather than from a reference.

    Why this matters more than it used to

    Symptom searches are exactly the queries where the result page has become least generous. SparkToro’s analysis of Similarweb’s US clickstream found 68.01% of Google searches ended without a click in early 2026, with AI Overviews appearing on over 20% of searches and cutting click-through by roughly 60% where they appear.

    A page that restates what sciatica is will lose that competition permanently. The summary above it already did that job. A page that says what happens in this clinic, how many visits it typically takes, what it costs and when you would refer out is answering something no overview can.

    Which is a reason to write fewer, more specific pages rather than more general ones.

    The local layer

    Condition pages compete in the ordinary results, where there is no distance constraint. The map pack is a separate contest ranking your profile, and they are two different competitions, so a page about sciatica is not a pack play and should not be sold as one.

    What it is, is the route by which somebody two suburbs beyond your pack radius finds you at all.

    The short version

    1. People search the condition, not the technique.
    2. Most clinics have five technique pages and no condition pages.
    3. Answer: what is this, is it serious, what does treatment involve, how many visits, what does it cost.
    4. Say when to see a doctor instead. Strongest trust sentence available.
    5. Six to ten conditions you actually treat. Not thirty.
    6. 68% of searches end without a click, so restating the textbook loses; clinic specifics win.
    7. Condition pages are organic, not map pack.

    If a site should be ranking and it isn’t, that’s the work I do. Chiropractic SEO covers the page set, the profile and the booking. If you’re not sure which of several plausible problems is costing you, that’s what an SEO audit is for.

  • “We’ll handle the insurance claim”: the page most roofers don’t have

    “We’ll handle the insurance claim”: the page most roofers don’t have

    TL;DR

    After a storm the homeowner’s problem is not the roof. It is the claim they have never filed and are frightened of getting wrong. Most roofing sites say they work with all insurance companies and move on. That sentence answers nothing. Build the page before the storm, because it cannot rank from a standing start.

    After a storm, the homeowner’s problem is not really the roof. It is the claim. They have never filed one, they do not know what their policy covers, they have been told the adjuster will lowball them, and they are frightened of doing something that voids the whole thing.

    Most roofing sites say “we work with all insurance companies” somewhere in a paragraph and move on. That sentence answers nothing, and the page that would answer it usually does not exist.

    Why it is worth a page of its own

    • People search it. Not “roof replacement”, but “does homeowners insurance cover roof damage”, “roof claim denied”, “insurance adjuster roof inspection”, “how to file a storm damage claim”. Those are informational queries with an enormous commercial event sitting behind them.
    • It is where the anxiety is. The roof is a known quantity. The claim is the unknown, and a business that removes the unknown wins the job before price is discussed.
    • Almost nobody has built it properly. Which is unusual. Most gaps in a competitive local market have been filled. This one persists because it is a page about insurance rather than about roofing, and it does not feel like the roofer’s job.

    What goes on it

    The sequence, in the order the homeowner will live it. What to do in the first 48 hours. Whether to file at all, including when the deductible makes it a bad idea. What the adjuster will look for. What happens if the claim is denied and what the appeal actually involves. What “we handle the claim” means in practice: what you do and what remains theirs.

    Then the honest boundaries, and this is the part that makes the page trustworthy rather than promotional: you are not a public adjuster, you cannot negotiate the claim on their behalf in most states, and there are rules about deductibles that vary by state and that some contractors ignore. Saying so plainly is the strongest trust signal on the page.

    The line to be careful about

    Storm-chasing outfits have made homeowners suspicious of exactly this pitch. “We’ll get your roof paid for” is what the truck with out-of-state plates says in the week after a hailstorm.

    So the page has to do the opposite of that. Specific, procedural, unexcited, with the limits stated. A local business explaining a process beats an out-of-town one promising an outcome, and the way you demonstrate you are the first is by writing something the second could not be bothered to.

    It also has to be a page you can service. Storm weeks produce call volume a small crew cannot absorb, and MIT’s study of 15,000 leads and 100,000 call attempts found the odds of qualifying a lead falling twenty-one-fold between a five and a thirty minute response. Generating demand you cannot answer is worse than not generating it.

    The seasonality of it

    This page needs to exist before the storm. Demand arrives in a spike measured in days and the page will not rank from a standing start. the start date decides the year and this is the clearest example of it in the trade.

    Written in the quiet season it is ready. Written the week after the hail it is competing with everyone else who had the same idea at the same time, from zero.

    What it does for the rest of the site

    Two things beyond its own traffic.

    It is the page other people link to. Nobody links to a roof replacement service page. A clear explanation of the claims process gets shared in local groups and referenced by neighbours, which is the only genuinely link-worthy asset most roofing sites can build.

    And it gives the trust conversation somewhere to happen before the phone rings. BrightLocal’s 2026 survey of 1,002 US consumers found 97% read reviews for local businesses and 31% will only use one rated 4.5 or above. Reviews carry most of that load, and a page demonstrating you understand the claim carries the part reviews cannot.

    The short version

    1. The homeowner’s problem is the claim, not the roof.
    2. The queries are informational with a large commercial event behind them.
    3. Almost nobody has built this page, because it does not feel like the roofer’s job.
    4. Write the sequence they will live, including when not to file.
    5. State the limits. Not a public adjuster, deductible rules vary. That is the trust signal.
    6. Be the opposite of the storm chaser: procedural and unexcited, not promissory.
    7. Build it before the storm. It cannot rank from a standing start.

    If a site should be ranking and it isn’t, that’s the work I do. Roofing and HVAC SEO covers the pages that get built and the ones that do not. If you’re not sure which of several plausible problems is costing you, that’s what an SEO audit is for.

  • Review recency in a trade that goes quiet for eight months

    Review recency in a trade that goes quiet for eight months

    TL;DR

    A roofer with a hundred and forty reviews, none of them from this year, is in a worse position than one with thirty from the last six months. Seventy-four per cent of people look for reviews written in the last three months. Seasonal trades get a permanent annual gap, and it lands in the researching months.

    A roofer with a hundred and forty reviews, none of them from this year, is in a worse position than one with thirty from the last six months. That is not intuitive and it is what the consumer data says.

    The number

    BrightLocal’s 2026 Local Consumer Review Survey, covering 1,002 US adults and published February 2026, found 74% look for reviews written in the last three months, and 32% want something from the last two weeks.

    The same survey found 97% read reviews for local businesses and 31% will only use one rated 4.5 or above, up from 17% a year earlier. So the bar is rising on rating and the window is narrow on recency, at the same time.

    Put those together and a review count is a stock while what people are actually reading is a flow. The total gets you into consideration. The timeline decides what happens next.

    Why seasonal trades are the worst case

    A restaurant collects reviews every week whether it tries or not. A roofer works in bursts, and if reviews are only requested during the busy months the profile has a visible eight-month gap in it. Every year, permanently, going back as far as anyone scrolls.

    Somebody comparing three roofers in February is looking at three timelines. Yours stops in September.

    Which is a prominence problem in Google’s terms as well as a persuasion one. Google names relevance, distance and prominence as the three local ranking factors, and adds that there is “no way to request or pay for a better local ranking.” Review activity is one of the few prominence inputs a small trade actually controls, and a seasonal business controls it for four months of the year.

    There is a second-order effect that matters more than the perception one: the gap arrives precisely when you are trying to book the next season’s work. The quiet months are when replacement and planned work gets researched, and that is when your profile looks least active.

    What I would actually do

    • Ask during the busy season, relentlessly, and bank nothing. The instinct is to space requests out. Do not. You cannot control when reviews arrive, only when you ask, and asking during the season is the only window you have.
    • Ask at the right moment. For a trade that is when the customer is standing in front of the finished work, not three days later by email. The completion conversation is the highest-yield moment available and most businesses waste it.
    • Do not ration or stagger requests. Some businesses hold reviews back to “spread them out”. You cannot bank a review that was never left, and the customer’s willingness decays within days.
    • Find the off-season reasons to be in touch. Maintenance visits, warranty checks, gutter clearing, service plans. These exist in most trades and they are the only genuine off-season review opportunity, which is an argument for offering them that has nothing to do with the revenue.
    • Reply to every review, including the old ones. A reply carries a date. A profile whose most recent activity is an owner response from last week reads as alive, even where the reviews are older.

    And be careful what you wish for on volume alone. A profile that generates more calls than the business can answer converts good visibility into bad opinions, which then arrive as reviews of their own.

    What I would not do

    Buy them. Beyond the obvious, a burst of reviews arriving in a month when you demonstrably were not working is the easiest possible pattern to spot.

    Chase the total. A firm fixated on getting to two hundred is optimising the number that matters least. Thirty recent beats a hundred and forty stale, and it costs less to maintain.

    Treat this as a ranking tactic. Review signals feed prominence, which is one of Google’s three local factors, but the honest case here is that 74% of your prospective customers are reading the dates. That argument does not need Google in it.

    The short version

    1. 74% look for reviews from the last three months. 32% want the last two weeks.
    2. Thirty recent beats a hundred and forty stale.
    3. Seasonal trades get a permanent annual gap, and it lands in the researching months.
    4. Ask during the season, at completion, in person.
    5. Never ration requests. You cannot bank one that was not left.
    6. Maintenance and warranty visits are the off-season opportunity.
    7. Reply to old reviews. A reply carries a fresh date.

    If a site should be ranking and it isn’t, that’s the work I do. Roofing and HVAC SEO covers the profile, the reviews and the season. If you’re not sure which of several plausible problems is costing you, that’s what an SEO audit is for.

  • HVAC has two buyers and one website

    HVAC has two buyers and one website

    TL;DR

    The system has failed and the house is cold. Or the system is fifteen years old and somebody is finally getting quotes. Two customers, two timescales, two definitions of a good website. Most HVAC sites are built for one and inherit the other by accident. Most of the fix is layout.

    The system has failed and the house is cold. Or the system is fifteen years old and somebody is finally getting quotes.

    Those are two different customers with two different searches, two different timescales and two different definitions of a good website. Most HVAC sites are built for one of them and inherit the other by accident.

    The emergency buyer

    Searching on a phone, now, with a specific fault or just “no heat”. They will call two or three companies from the map pack and whoever picks up gets the work.

    What this buyer needs from the site is almost embarrassingly simple: a phone number that works with a thumb, a clear statement that you cover their area, and confirmation that you can come today. Everything else is in the way.

    They will not read your accreditations. They will not compare financing. The decision is made in under two minutes and it is mostly decided by who answers. MIT’s study of 15,000 leads and 100,000 call attempts found a twenty-one-fold drop in the odds of qualifying a prospect between a five-minute and a thirty-minute response, and an emergency caller is working down a list rather than waiting for you.

    The replacement buyer

    Researching over days or weeks, often on a laptop, frequently with a spouse. Comparing three quotes. Reading about efficiency ratings, warranties, rebates and finance.

    This buyer wants depth, and they want the money conversation answered before they call. What does a system like mine cost. What finance is available. What is the rebate situation this year. How long does an install take and will I have heat that night.

    They also read reviews properly. BrightLocal’s 2026 Local Consumer Review Survey found 97% read reviews for local businesses and 31% will only use one rated 4.5 or above. A five-thousand-dollar decision is exactly where that behaviour concentrates.

    Why one site struggles to serve both

    The two buyers want opposite things from the same real estate.

    Strip the home page for the emergency caller and the replacement buyer finds nothing to evaluate. Load it with efficiency comparisons and finance calculators and the emergency caller cannot find the phone number.

    The commonest failure I see is a site built entirely for emergencies, with a big phone number, “24/7” and little else, by a business whose actual margin is in replacements. It is optimised for the calls that pay least.

    How I would split it

    • Separate the paths from the first screen. Two clear routes: something is broken now, or I am planning a replacement. This is a layout decision, not a content one, and it is most of the fix.
    • Emergency pages: short, fast, phone-first. One per fault type where volume justifies it: no heat, no cooling, water leaking from the unit. Deadline-style urgency at the top.
    • Replacement pages: long, and unusually so. Sizing, efficiency, what the install involves, and the money answered plainly. A financing page is an SEO asset, not a sales page. People search for it.
    • Different measures of success. Emergency is calls answered. Replacement is quotes issued. Reporting them as one number tells you nothing about either.

    The seasonal overlay

    These two peak at different moments. Emergencies spike on the first cold night and the first heatwave. Replacements cluster in the shoulder seasons, when nothing is urgent and people are getting organised.

    Which means the content for each should be ready at different times, and and the start date decides whether it is. Replacement content built in autumn is late for the autumn shoulder and early for the spring one.

    The short version

    1. Two buyers, opposite requirements, one site.
    2. Emergency: phone, area, today. Everything else is in the way.
    3. Response speed decides it: a 21-fold drop between five and thirty minutes.
    4. Replacement: depth, and the money answered before they call.
    5. 97% read reviews; 31% require 4.5+. That concentrates on the big-ticket decision.
    6. Split the paths from the first screen. Most of the fix is layout.
    7. Measure calls answered and quotes issued separately.

    If a site should be ranking and it isn’t, that’s the work I do. Roofing and HVAC SEO covers both buyers and the phone behind them. If you’re not sure which of several plausible problems is costing you, that’s what an SEO audit is for.