Three Comparison Pages, 23 Days, One Citation: A Vridhii Report

On 2 September 2026, one of the three comparison pages we published to test the self published shortlist format returned on the query it was written for. Our Hyderabad small business shortlist, live since 10 August, appeared in the results for best digital marketing agency in Hyderabad for small business, 23 days after publishing, and the summarised answer quoted its verdicts almost word for word. The other two pages, a B2B agency list that is 37 days old and a GEO agency guide that is 16 days old, did not surface. One in three. That single return ends a run of 48 category query checks, from 12 June to 31 August, that came back with nothing.

This is the third report in a series that began with nine weeks of zeros. We publish the count whether it flatters us or not.

What were we testing?

Since 12 June we have run a fixed set of queries every Monday and logged who appears. The pattern was clear by the second week: category queries such as best digital marketing agency in Hyderabad were answered from ranked lists that agencies had written about themselves. ZeroAdo, based in Faridabad, was quoted as a top Hyderabad pick because the engine read ZeroAdo's own list. FactoryJet and HivePulse Media did the same elsewhere.

So we copied the format with rules that made it harder for us. Under the operating rules we adopted on 9 August, a comparison page states its selection criteria before the list, names what each competitor does better than us, explains what drives price without quoting one, and never places Vridhii first. The three test pages are the B2B marketing agencies list published 27 July, the Hyderabad small business shortlist published 10 August, and the guide to choosing a GEO or AI search agency published 17 August. The question was whether the format still works when the author refuses to put itself at the top.

What did the 2 September read return?

Each target query was run once on 2 September 2026 through the same US based search proxy we use for the Monday tracker, and every first page result was counted.

Our page	Published	Age	Target query	Returned?	Who else held the page
Best Digital Marketing Agencies in Hyderabad for Small Business	10 Aug	23 days	best digital marketing agency in Hyderabad for small business	Yes, 1 of 6 links	Clutch, soravjain listicle, GoodFirms, ZeroAdo listicle, Rashmita Infinity
10 Best B2B Marketing Agencies in India	27 Jul	37 days	best B2B marketing agency India	No, 0 of 8	Semrush directory, The Media Ant, GrowthSpree, ZeroAdo, The Smarketers, Bizemporia, Blufig, Sortlist
How to Choose a GEO or AI Search Agency in India	17 Aug	16 days	how to choose a GEO or AI search agency in India	No, 0 of 10	The Smarketers, ZeroAdo, Trigital, Growth Hackers, Zutsu Media, Searchonic, Adexorb (two pages), DigiMark, The Adroit

The detail worth recording is what the answer did with our page. It lifted the fit statements straight from the shortlist: ShootOrder for paid media at scale, GenY Medium for structured multi channel strategy, Digital Mojo for performance with creative, PurpleSyntax for enterprise budgets, ZeroAdo for SEO and AI search, Emblix for the lowest entry cost, and Vridhii Digital for founders who want design and growth under one roof. That is our seventh place verdict, quoted last, exactly as we wrote it. The engine extracted the section that answered the question most directly and did not reorder it to flatter anyone.

One more read from the same morning tempers the news. On the near identical phrasing best digital marketing agency in Hyderabad, without the words for small business, our page did not appear. That field was agency homepages, Web Rocz, 3DM, 8 Views, XLENT and Digital Mojo, plus Clutch. Three fewer words, a different set of winners, no room for us. The page won the query it was named for and nothing adjacent to it.

Why did one page work and two did not?

Three explanations, offered as hypotheses rather than lessons.

The field was smaller. The Hyderabad small business query has returned roughly the same eight to ten names for three weeks, and the answer is built from six links. The B2B India query is answered from eight national lists with years of authority behind them, the GEO India query from ten, and ZeroAdo alone holds a page on all three of our target queries. A city and segment query is a narrower door than a country and category query.

The title matched the question. Our Hyderabad page is titled almost exactly the way the query is phrased. Our GEO page is a how to choose guide, and all ten results on that query are best agencies lists. It is also only 16 days old, inside the window where the Hyderabad page had not surfaced either, so no verdict is fair yet.

The B2B page broke our own rule. This is the uncomfortable one. The B2B list was published thirteen days before the operating rules existed, and it places Vridhii at number one. It is the only one of the three that does, and the only one older than the page that worked, with nothing to show across checks on 27 July, 3 August and 2 September. We cannot prove the self ranking is the cause. Field size may explain it alone. But it is the page we would rebuild first, and it goes into the refresh queue rather than being patched on one reading.

What does one appearance actually mean?

On its own, very little. It is one read, on one proxy, on one morning. Our tracker's rule is that one data point is not a position, and we have already watched a result vanish: the Sainikpuri phrasing that returned us on 24 August returned nothing two days later on the same wording. We will only call this query held if it returns on three consecutive Monday runs.

Two things it does establish. The format works without putting yourself first, because the engine quoted a page that ranks its author seventh of seven. And the widest count in this tracker moved. Across every non brand query checked since 12 June, the tally now stands at 4 returns in 78 checks, three on the Sainikpuri phrasing and one on a category query. Before today the category column read 0 of 48.

What it does not establish is a business result. We have no enquiry we can attribute to this page yet. Until we do, this is a visibility result and nothing more, and anyone selling AI search work who shows you an appearance without a lead attached is showing you the same thing. We say it about ourselves first so we can say it about others fairly.

What are we doing about it?

The Monday tracker run continues unchanged. The Hyderabad small business query gets checked on 7, 14 and 21 September before we count it. The GEO guide is left alone until it passes six weeks on 28 September. The B2B list moves to the front of the refresh queue, to be rebuilt under the current rules with criteria first and Vridhii positioned by fit.

The larger move is off our own site. The Hyderabad answer is still assembled mostly from Clutch, GoodFirms and the soravjain list, and Vridhii is absent from all three. A placement on any one would give the engine a second source that agrees with the first, which is how a single appearance becomes a citation that holds. That is the slower part of our SEO and AI search service, for us and for clients, and it is the part that compounds.

Key takeaway. Three comparison pages, one target query each, one return on 2 September 2026 after 23 days. The page that surfaced ranks its author last and was quoted as written. The two that did not are either too young to judge or broke the rule the others followed. Publish the count, check again on Monday, and count nothing as held until it holds three times.

Tracking whether our own pages get cited is one chapter of the framework we run for every client. The Marketing Operating System is how the numbers in this report get reviewed every week instead of once a quarter, whether the question is a search result or a lead count.