Aglocom SEO Autopilot
Managed search visibility, priced like software.
We get your website showing up when people look for what you sell — on Google, and when they ask ChatGPT or Perplexity. Every month you get two scores that show what moved, measured the same way every time so you can actually compare them.
₹20,000/month. One site. No setup fee, no lock-in.
Performance carries 10% of SEO Health, so its fall is already inside the top bar. It is shown separately here.
18→to66
AI Visibility, out of 100, in five weeks
+20.9
SEO Health points out of 100, same window
5.1→to3.1
Avg position, "best mpc colleges in hyderabad"
Scores are from one engagement, measured 15 July to 19 August 2026 on the rubric below. The position figure is a second client, Search Console 28-day averages for the windows ending 6 April and 25 April 2026 — a keyword that was unranked before the engagement. Both are detailed further down, including the category that went backwards.
Will you rank in seven days? No.
New pages take two to four weeks just to register with Google, and competitive positions move over months. Anyone promising otherwise is guessing. What the first week buys is a fixed foundation and a running engine, both measured — and from month two we report movement against the number we took before touching anything.
The gap
Configuration problems, mostly.
Google still sends traffic. But ChatGPT, Perplexity and AI Overviews increasingly answer the question first, and they can only cite what they are allowed to fetch and able to parse.
On the engagement we can show you in full, four things accounted for almost all of the missing visibility.
The firewall was turning the AI crawlers away
The site's WAFA firewall in front of a website that blocks visitors it judges suspicious. It cannot always tell a useful robot from a harmful one. returned 403The server's way of saying “you are not allowed in”. A person sees the page; whoever is blocked sees a wall. to both search-index crawlersAutomated readers. Google has one, and so do ChatGPT, Perplexity and Claude. They fetch pages so their systems can quote them. — OpenAI's and Anthropic's — on every page, so it could not enter either citation corpusThe pool of pages an AI assistant is willing to quote when it answers a question. Outside it, you cannot be recommended — however good you are.. The trigger was a generic bad-bot rule matching the substring "searchbot": the two blocked agents were the two whose names contain it, and eleven other crawlers passed fine. The client's monitoring tool reported all sixteen AI crawlers as accessible, because it read the robots rules instead of sending requests.
There was no structured dataHidden labels in a page telling software what each thing is — a price, an author, an answer to a question. Without them a machine has to guess.
None at all, so no rich results and no clean fact for an answer engine to lift with attribution.
No llms.txtA plain text file, read by AI assistants, that says what a business does and which pages hold the answers. New, and increasingly consulted.
The file that tells AI systems what the business is and which pages carry the answers.
One template was breaking every page
Twenty-one of 39 pages had no H1 because the post template rendered the title as an H2. Nine more had none for unrelated reasons. We found it; the client changed one widget, which fixed all 21 at once and made every future post right by default.
Month one
Seven working days from access to a running system.
You grant access on day zero. By day seven the foundation is repaired, the plugin is live, the engine is publishing, and you have a before-and-after scorecard.
Access and scoping
Search Console and Analytics access, a WordPress application password, and a half-hour call to record what you sell and what we may never claim for you.
Baseline audit
We fetch and parse every URL in your sitemap. You get two scores and a gap registerThe running list of what is wrong, ranked by severity, each item saying who owns it — us or you., each finding marked as ours to fix or yours to decide. The baseline freezes here.
Technical repairs
The parts a search engine reads before it reads your words: what appears as your headline and blurb in results, the hidden labels that tell software what each page is, and the settings that quietly keep pages out of listings. In full: titles, meta descriptions, Open Graph, JSON-LD, breadcrumbs, canonicals, sitemap cleanup, security headers, XML-RPC closed.
The AI layer
We check, one assistant at a time, that each can actually fetch your pages — not that your settings say it may, but that your server really answers it. Then we publish the file they read to understand your business, and the labels they quote from. If a firewall is turning them away we bring you the trade-off and you decide, which is the step most likely to run past the week.
Research and calendar
Live search volume, competitors' ranking keywords, and the queries you already collect impressions for and lose. Out of it comes a dated 90-day calendar, one job per post.
Publish and re-score
The first articles go live in your design language, under your byline, links and schema attached. Then we re-run the audit and hand you the delta.
What seven days cannot do
It cannot move your rankings. New pages take two to four weeks to register in Search Console at all. Ranking movement is reported at 60 to 90 days against the frozen baseline.
We also will not predict how far your scores move that week. It depends on what the baseline finds and how fast your side answers.
The numbers
One engagement, in full.
A US healthcare analytics company on WordPress, behind a firewall. Baseline frozen 15 July 2026, re-measured 18–19 August by fetching all 39 sitemap URLs. Five weeks, some of it spent waiting on decisions only the client could make.
Score A · SEO Health
Seven scoring categories. Each row gives the category, its weight in the score, the frozen baseline of 15 July 2026, and the 19 August re-measurement.Score B · AI Visibility
Six checks over the same window. Each row gives the check, its weight in the score, the frozen baseline of 15 July 2026, and the 19 August re-measurement.The tick is the frozen baseline; the bar is the re-measurement. Twelve of thirteen rose. Performance fell 48 to 28 and is shown in amber; AI Review Signals did not move. Weighted, these rows are the two headline scores — nothing is averaged and nothing is left out.
One template change, twenty-one pages
We found the post title widget set to H2 instead of H1. The client changed it in a day, and pages with exactly one H1 went from 9 of 39 to 30 — every future post correct by default. That one change re-scored two categories on its own: On-page SEO +21 and AI Content Readability +18. Finding it is our job; shipping it was theirs.
What did not move, and why
AI Review Signals scored 18 at baseline and 18 at re-measure. It is gated on business decisions — review volume and how you answer them — not on anything we can ship. Performance fell 48 to 28 in the same window: the client removed the WAF so the AI crawlers could reach the site, and it took the CDN edge cache with it. Median time-to-first-byte went 0.42s to 2.39s. A scorecard that only goes up is a brochure.
A second market
Nine Education, a Hyderabad coaching institute. During the engagement best mpc colleges in hyderabad went from unranked to average position 5.1, then 3.1; mpc colleges in hyderabad moved 3.5 to 2.5. Search Console 28-day averages for the windows ending 6 and 25 April 2026, from pulls dated 14 and 27 April. Terms already ranking before we arrived are not claimed.
Re-pulled 24 August, and it is not all up. Publishing stopped in May. Over the 28 days to 21 August those two terms sit at 4.8 and 5.4 — both below where the engagement left them. Non-branded page-one keywords at 30 impressions or more went the other way: 136 in April, 240 now. We are showing you the ones that fell because a case study you cannot check is worth nothing. The case study covers it in full.
The scorecard
Two scores, on weights that stop moving once we start.
Both scores are out of 100, and the way they are worked out is fixed before we start and never changes while we work. That is the whole point: this month's number can be compared with next month's, and with the one we took before touching anything.
We learned it expensively. On an early engagement the scoring changed between the first audit and the second, so the headline number described a change in the scoring rather than any work, and the two audits were never comparable.
The rubricOur scoring sheet: fixed categories, fixed weights, and a written description for every band. Frozen on day one so this month’s number can be compared with next month’s. is now versioned and frozen at your baseline. Categories are never added, dropped or re-weighted mid-engagement, and each has written scoring bands.
Every figure comes from fetching your live pages. Where a tracker disagrees with our crawl, the crawl is the score of record.
Score A — SEO Health
| Category | Weight | Weight, drawn |
|---|---|---|
| Technical SEO | 25% | |
| Content quality & E-E-A-TExperience, Expertise, Authoritativeness, Trust — what Google looks for to decide whether a page was written by someone who actually knows the subject. | 25% | |
| On-page SEO | 20% | |
| Schema / structured data | 10% | |
| Performance (Core Web VitalsGoogle’s three measures of how quickly and smoothly a page loads for a real visitor.) | 10% | |
| Images | 5% | |
| Internal linking & architecture | 5% |
Score B — AI Visibility
| Check | Weight | Weight, drawn |
|---|---|---|
| AI schema & structured data | 25% | |
| AI crawler access | 20% | |
| llms.txt | 15% | |
| AI content readability | 15% | |
| AI content gap | 15% | |
| AI review signals | 10% |
Month two onward
What runs without you.
Ten new pages a month, published for you
Ten posts a month, one every third day, each written to a query from your calendar with links and schema attached before it goes live.
Nothing breaks quietly
Every week we re-check the things that break quietly: a page accidentally hidden from search, the files AI assistants read no longer being served, links gone dead, publishing stalled. Anything failing raises an alert rather than sitting in a log.
A fresh number every month
The full crawl again, scored on the same frozen rubric, with the delta against last month and your original baseline.
What actually moved in Google
Live Search Console positions for your tracked keywords: what moved, what slipped, and which new queries you started appearing for.
We chase whatever is working
Queries you are already collecting impressions for get promoted. Posts that stopped earning their slot get cut.
The technical side stays current
Schema kept current as the site changes, new page types added to the meta layer, and up to six gap-register findings shipped a month. Anything beyond that is queued and prioritised with you.
How it reaches you
A plugin on your site, a report in your inbox.
The plugin does the technical work on your site, automatically and continuously: the headline and blurb each page shows in results, the hidden labels that let software quote you accurately, the file AI assistants read to understand your business, and the rules that let their crawlers in. Your scorecard arrives as a report — both scores against the frozen baseline, the open gap register, the publishing queue and crawler-access status — after the day-seven audit and monthly after that.
The same scorecard also renders inside your WordPress admin, included when your site is on WordPress. It is new: it works, it is tested, and no client has run it for a full month yet. You would be among the first, and we would rather say that than let you find out.
We are fixing these
- Restore edge caching Performance (CWV) · in progress
- Ten pages render zero H1 On-Page SEO · open
These need your decision
- Publish third-party review profiles AI Review Signals · waiting on you
- Clear two customer case studies for publication AI Content Gap · waiting on you
Scored, and not fixable by us. We do not count them as work in progress and we will not forecast a score movement on them.
Rubric 1.0 · frozen at 2026-03-01 · both scores measured 2026-04-12 · Example Co is a fictional client, and these are its numbers, not anyone else's
01 In your WordPress
02 Day one · before there is a score
How it is wired
- Data in: Search Console, Analytics and Ads, keyword and competitor data, and a live crawl of every URL in your sitemap.
- The engine, on our side: audit and scoring on the frozen rubric, the gap register, a 90-day calendar built from real demand, then draft, review, schema and publish, with a weekly monitor over all of it.
- Your site: posts arrive over the WordPress REST API, and the plugin applies meta, schema, llms.txt, robots rules, headers and sitemap hygiene.
- What you do: grant access once, approve the calendar, read the monthly report.
Day zero
What we need from you, once.
Every grant goes to the dedicated Aglocom service account named in your kickoff email, never to an individual's personal account, so access survives staff changes on both sides.
Under an hour on day zero. After that, about an hour a month to read the report and approve the next calendar.
Required
- Google Search Console: full user for our service account
- Google Analytics 4: viewer access
- WordPress: an application passwordA separate password WordPress creates for one tool, which you can revoke on its own without changing your own login., editor level or above
- The ability to install one plugin
- The name and title of whoever signs off as author
- Half an hour on a call to agree what we may and may not claim for you
Helpful, not blocking
- Google Ads read access, for real search volume in your market
- Firewall or CDN dashboard access, if a WAF sits in front of the site
- Your three main competitors, named
- Any previous audits you have been given
Fit
Who this works for.
A good fit
- WordPress sites of roughly 20 to 500 pages
- Companies whose buyers research before they buy, increasingly by asking an AI
- Teams with no in-house SEO, or one person carrying it alongside other jobs
- Anyone who has paid for SEO before and never received a number they could check
A bad fit
- Sites that need traffic this month
- Businesses with no site, or one being rebuilt in the next 60 days
- Anyone who wants a volume of posts as the deliverable
- Anyone expecting guaranteed positions
Pricing
₹20,000 a month. Month one is the proof.
No setup fee and no minimum term. Month one is billed and delivers the audit, the fixes, the plugin and the first posts. If the day-seven scorecard does not convince you, there is no month two.
₹20,000 / month
One site. Everything below, running continuously.
- Baseline audit on the frozen two-score rubric
- Plugin installed, configured and maintained
- Technical, schema and AI-access fixes — up to six findings a month
- 10 published posts a month
- Weekly automated health check
- Monthly re-audit, score delta and search scorecard
- Your scorecard as a report, after day seven and monthly
- Extra content volume beyond ten posts a month
- Additional sites or subdomains, each with its own baseline, calendar and onboarding week
- Sites above roughly 500 URLs, where the monthly full crawl is a real cost
- Google Business Profile management. Local "near me" demand is real and content alone cannot capture it, but we do not sell that service today
- Paid advertising. For education clients that is our Admissions Growth System; outside education we do not sell it
- Landing page design or site rebuilds
- WhatsApp, email or CRM nurture
- Bought backlinks and link schemes, at any price
Questions
Common questions.
Will I rank in seven days?
No. New pages typically take two to four weeks just to register in Search Console, and competitive positions move over months. Week one buys a fixed foundation and a running engine, both measured. Ranking movement is a month two and three conversation, reported against the baseline we froze before touching anything.
Is the content AI-generated?
It is AI-drafted and human-reviewed. Every client gets a constitution file first: what you sell, what you do not sell, what may never be claimed, who the author is, and which numbers are allowed in writing with their source. Nothing publishes that contradicts it.
What happens if I stop?
Every published post stays on your domain. The plugin stays installed in its last shipped version and keeps serving the fixes while it is active, but it stops receiving updates and the reports stop. If you later deactivate it, the meta and schema layer it was serving goes with it.
Why is this ₹20,000 when agencies quote much more?
Because most of the work is carried by a system rather than by headcount. Auditing, writing, publishing and monitoring all run as software, with a human reviewing the output.
Do you work outside education?
Yes. The engine has been built and run for both an Indian coaching institute and a US healthcare analytics company. The two-score rubric itself was built on the healthcare engagement and is vertical-agnostic by design; the coaching engagement is where we learned what happens when a rubric is not frozen. The vertical changes the constitution file and the keyword research. If you are a school or coaching institute wanting search and ads and nurture as one funnel, that is our Admissions Growth System instead.
What if my site is not WordPress?
The audit, scoring, research, content and reporting all work anywhere. The plugin does not. On other platforms the fixes arrive as implementation tickets for your developer, and the reporting is identical. Tell us your stack.
Next step
Find out what your two scores are.
We will run the baseline audit on your site and show you both scores, the gap register, and what week one would fix.