Over 300 AI citations in a month: Ramp’s two company-size comparison pages

Two people looking for the “best software” may need very different answers when one runs a small business and the other works at a large enterprise. Ramp addressed that distinction.

The company reportedly earned more than 300 AI citations in one month across two company-size-specific accounts payable pages. Profound’s case study

Accounts payable management covers handling invoices and paying suppliers. The practical AI search optimization lesson is to organize explanations around the buyer’s conditions.

Ramp’s problem: low visibility for accounts payable questions

Ramp identified weak AI visibility in the accounts payable category. Being known as a company did not mean appearing equally across every product topic. Profound’s case study

Illustration: Brand recognition does not necessarily mean appearing as an option in a product category.

Check branded explanations and unbranded comparisons separately.

An AI may correctly answer “What does our company do?” while omitting it from “Which invoice tool works for a small team?” Check whether a page actually answers the latter question.

Different company sizes led to separate comparison pages

Ramp created separate accounts payable software pages for small businesses and large organizations. Profound’s case study

Illustration: Small and large businesses may need different operational details when choosing software.

The current small-business guide and large-business guide explicitly identify their audiences. Their current versions may differ from those used during the reported measurement period.

The editorial lesson is to reflect audience differences in comparison criteria, not just in the title.

As an illustration, a small team may care about operating without a dedicated administrator, while an enterprise may need multiple approvers and system integrations. Select actual criteria from your own customer evidence.

The result: more than 300 citations across two pages, with a visibility-chart caveat

The reported outcome is more than 300 citations across two pages in one month. A citation is not a visit or a signup. Profound’s case study

Illustration: Keep page citations separate from visibility changes with unresolved measurement differences.

For visibility, the source’s prose describes a change from 3.2% to 22.2%, while the chart displays a different starting value and period. Because their relationship remains unclear, this article does not use a visibility multiplier in its headline.

The vendor-published story does not isolate page separation as the cause or establish that another company will reproduce the result.

Try it yourself: separate buying questions by customer size or role

Choose two customer groups with different purchasing conditions and compare their questions. This is our proposed exercise.

Illustration: Use sections for shared answers and consider separate pages for substantially different needs.

Illustrative customer	Decision condition	Information to explain
Small team	No dedicated administrator	Setup, daily tasks, and support
Multi-location business	Branch-specific approvals	Permissions, approval sequence, and consolidated reporting

Use sales and support records to identify actual questions. Separate those with a common answer from those whose answer depends on the customer’s situation.

When comparing other products, check official information under the same criteria and retain the verification date. Include unsupported conditions as well as strengths.

Decide whether separate pages or sections will answer the questions better

If most answers are shared, audience-specific sections on one page may be sufficient. Consider separate pages when implementation conditions and comparison criteria need substantially different explanations.

After publication, keep audience-specific questions consistent and observe which URLs are cited. If a small-business question returns an enterprise page, review the body’s target audience and internal links as well as the title.

The goal is an answer that fits the reader’s situation, not a larger page count.