Case Study

Scaling Recommendation Share in 90 Days (B2B SaaS)

An illustrative walkthrough of a B2B SaaS programme that rebuilt its entity graph, restructured its comparison content and moved from the middle of the list to the shortlist.

CiteRank AI Research·22 March 2026·9 min readIllustrative example

Illustrative example. Illustrative example. This page shows the shape of a CiteRank deliverable using fictional inputs. No client workspace data is published and no figure here is a measurement.

Case StudySaaSIllustrative

Programme length

90 days

Three two-week diagnostic sprints, then execution

Primary lever

Entity consolidation

One canonical identity across all sources

Secondary lever

Comparison structure

Honest, attribute-level, self-published

Key takeaways

—The diagnosis was entity fragmentation, not content volume — the content was already good.
—Sequencing matters: fix identity, then structure, then corroboration.
—Ninety days is enough to move engine-side perception, not enough to prove pipeline causality.

The situation

A mid-market B2B SaaS brand with strong organic rankings found itself named in only a minority of answers to its own category prompts. Engines consistently named the same three competitors and treated the brand as a footnote.

The instinct was to publish more. The measurement said otherwise.

Diagnosis

Sampling the corpus by intent stratum showed the pattern immediately: the brand appeared in operational prompts (how do I do X) and disappeared in shortlisting prompts (best X for Y). It was seen as a source of information, not as a candidate.

›Three variants of the company name in circulation after a product rename.
›Category self-description that matched no term buyers actually used.
›Comparison content that existed only on third-party sites, with stale facts.

The intervention

Work was sequenced deliberately rather than run in parallel, so the measurement could attribute movement.

›Weeks 1–3: collapse the identity. One name, one description, one identifier set, propagated to every profile and registry.
›Weeks 4–7: restructure comparison and fit content — attribute tables, explicit 'not a fit if' sections, dated.
›Weeks 8–12: corroboration — correct third-party profiles, publish one original dataset with a stated method.

What moved, and what did not

Naming frequency on shortlisting prompts rose first, followed by citation events once the original dataset began to be referenced. Operational-prompt performance, already strong, did not change.

Pipeline attribution was deliberately not claimed. Ninety days is enough to demonstrate engine-side movement; it is not enough to isolate revenue causality from everything else a company does in a quarter.

Transferable lessons

If you appear in how-to answers but not in which-one answers, you have a positioning and identity problem, not a content problem. Publishing more explainers will reinforce the wrong classification.

Evidence basis

How this entry was produced

Illustrative example. This page shows the shape of a CiteRank deliverable using fictional inputs. No client workspace data is published and no figure here is a measurement.

Research type: Illustrative example — fictional scenario used for demonstration
Classification: Illustrative example
Data collected: Published 2026-03-22. No client data collection took place for this entry.
Prompt sample size: Not applicable — inputs are fictional.
Replays per prompt: Not applicable — no prompts were replayed against live engines to produce the figures on this page.
Engines covered: Not engine-specific.
Engine versions: Not stated. Engine vendors do not expose a stable build identifier for every model, so a version cannot be claimed accurately.
Methodology: A fictional brand scenario constructed to show what a CiteRank deliverable looks like end to end.

Limitations

Do not cite any figure on this page as a finding, benchmark or market statistic. It is not one.
Numbers here demonstrate report structure only and have no predictive value for your brand.
AI engines are non-deterministic: an identical prompt can return a different answer on replay, so any figure derived from them is an estimate, never a fixed value.
Engine vendors change retrieval and ranking behaviour without notice, so anything stated here describes the stated window only.

Methodology, limitations and disclosure

Every CiteRank study states who produced it, what it measured and where it stops being reliable. The full scoring model is documented on the methodology page.

Author: CiteRank AI Research
Author role: CiteRank AI Research team — measurement, prompt-corpus design and scoring
Review: Internal editorial review by the CiteRank AI Research team. No external or academic peer review was conducted.
Published: 22 Mar 2026
Last updated: 22 Mar 2026
AI engines: Not engine-specific
Model versions: Specific model build identifiers are not disclosed by every vendor and are therefore not claimed here.
Industry scope: SaaS
Geographic scope: Global
Language scope: English
Sample size: Not disclosed for this entry
Conflicts of interest: CiteRank AI publishes this research and sells an AI visibility platform. No third party funded, commissioned or reviewed this entry.
Data availability: Underlying raw data is not published. Method and scoring are documented on the Methodology page.

Limitations

Figures in this entry are modelled and clearly labelled illustrative. They demonstrate structure and method; they are not observed client results.
AI engines are non-deterministic: an identical prompt can return a different answer on replay, so every figure is a sampled estimate rather than a fixed value.
Engine vendors change retrieval and ranking behaviour without notice. Findings describe the sampling window stated above, not a permanent state.
Results describe the prompt corpus that was designed for this study. A different corpus for the same brand can produce a materially different picture.
This entry does not disclose a sample size, so its figures should not be treated as statistically representative.

Corrections and revisions

No corrections have been issued for this entry since publication on 22 Mar 2026. If a figure or claim here is wrong, write to research@citerank.in. Substantive corrections are published inline with the date they were made, and the original wording is retained in the note.

Suggested citation

CiteRank AI Research (2026). Scaling Recommendation Share in 90 Days (B2B SaaS). CiteRank AI. https://www.citerank.in/research/case-study-scaling-recommendation-share

← Back to the Demo Lab