Case studySmall-business CRM%22%20d%3D%22m0%200%2016%2011h61%2061%2061%2061%2060L47%2037h61%2061%2060%2061L16%2063h61%2061%2061%2061%2060L47%2089h61%2061%2060%2061L16%20115h61%2061%2061%2061%2060L47%20141h61%2061%2060%2061L16%20166h61%2061%2061%2061%2060L47%20192h61%2061%2060%2061L16%20218h61%2061%2061%2061%2060z%22%2F%3E%3C%2Fsvg%3E)
Name under NDA
US$440K in additional subscription revenue for a US CRM
- US$440K
- 1,600
- +29%
Who the client is
Business and buyers. This US B2B CRM serves small businesses and sales teams that want an approachable system without a dedicated administrator. It combines contact management, sales pipelines and email marketing with guidance on the next action for each lead.
Opening scale. The engagement began with 5,000 paying companies, 20,000 seats and approximately US$10.3 million in core CRM annual recurring revenue. The marketing website had 1,200 public pages. Our work targeted US buyers searching in English; website and subscription totals cover all markets.
Plans and support. Three plans charge per seat. Marketing contact charges and optional services are separate. The product offers a fourteen-day trial without a card, live support and onboarding.
Starter
US$19per seat a month
US$13 a month billed annually
Contact management, email/calendar sync, shared tasks and basic reporting
Professional
US$49per seat a month
US$42 a month billed annually
Sales automation and multiple pipelines
Business
US$67per seat a month
US$59 a month billed annually
Team and territory controls, plus audit logs
What the client came to us with
Commercial goals. The client wanted more qualified trials and paid subscriptions from organic search. Buyers needed clearer answers about adoption effort, reliable follow-ups, connected sales and marketing work, and the total cost for their team.
AI recommendations. The team also wanted more citations and favorable recommendations that explained the product’s fit accurately. Answers needed to distinguish its plans, integrations and limits so suitable buyers could make an informed choice.
Results the client got
Additional revenue. From January 15 through August 31, organic-acquired customers generated an estimated US$980,000 at their original subscription terms, after a 2% credit provision. The chart below compares it with the estimate for continuing the client’s previous approach to search and content.
- US$540Kestimated with the previous approach
- US$108Kcovered our fees
- US$332Kgained on top
Our fees were EUR96,000, about US$108,000.
Paying customers. By August 31, about 3,000 companies acquired through organic search during the engagement were paying customers. Their initial subscriptions represented about US$330,000 per month, including an estimated US$170,000 above the US$160,000 expected with the previous approach.
- Organic traffic. Website visits from organic search increased about 29% for January 15-August 31 compared with the same dates in the previous year.
- Qualified trials. Organic search brought about 20,000 trial accounts with a verified business, a supported use case and a supplied team size during the reporting period.
- AI citations. We monitored the same 120 buying questions across ChatGPT, Perplexity, Gemini and Claude. Of 1,440 answers collected in each measurement, 86 linked to the client’s website on January 15-17, compared with 216 on August 30-September 1.
- Recommendations with checked claims. Favorable recommendations with accurate checked claims and a specific fit reason rose from 36 to 190 within those answer sets.
January 2026September 2026
Answers linking to the client websiteOf 1,440 sampled answers
Share within selected brandsClient appearances among all appearances of the client and four competitors: 144 of 1,440, then 288 of 1,600
Favorable sentimentAmong answers mentioning the client
Product precisionCorrect claims among assessed claims
Recommended with checked claims and a fit reasonFavorable, accurate and with a concrete reason the CRM suits the buyer
The situation before the engagement
Business context
Established competition. The client competes with better-known CRM brands. Its marketing, product and support teams already run a mature acquisition operation. Buyers compare ease of adoption, follow-up automation and team costs, so general descriptions of contact management leave important purchase questions unanswered.
Website and content
Existing content. The site contained 600 articles, 280 commercial pages, 100 glossary entries, 130 integration/template pages and 90 company or other pages. General sales education was extensive, but important buying and setup information was disconnected:
- The automation page described features without identifying the required plan.
- Pricing information did not clearly separate unlimited CRM records from chargeable marketing contacts.
- Migration instructions omitted custom-field mapping and duplicate handling.
- Trial emails sent users to a generic help index instead of relevant setup and evaluation guidance.
Buyer uncertainty. Small teams struggled to establish their total cost, select the right tier and recreate their existing sales process. The product already supported the required workflows; buyers needed clearer explanations and instructions.
Technical obstacles. Important comparison text was missing from the initial HTML, and some integration pages identified the wrong URL for indexing. Repeated titles and vague descriptions gave buyers little reason to click. Structured data and internal links also needed to reflect the product and connect related buying information.
Traffic sources
Website traffic. The public marketing website received 1.65 million visits in the preceding year: 720,000 from organic search, 480,000 from direct visits and referrals, 300,000 from paid campaigns and 150,000 from other sources. These were GA4 sessions, not unique people. Usage inside the CRM application was excluded.
- 720,000Organic search
- 480,000Direct visits and referrals
- 300,000Paid campaigns
- Other sources (150,000)
Search presence
Starting visibility. Organic traffic averaged 60,000 monthly sessions, with half landing on commercial pages. Only 38 of 200 tracked buyer queries ranked in Google’s top ten. AI answers often listed the product generically, omitted its sales and marketing fit, or confused tier features and prices.
38 in the top ten at kickoff
82 on September 1
- Google top three
- Positions four to ten
- Below the first page
One square per tracked buyer query.
Previous approach
Disconnected decisions. Articles were commissioned individually and product approval arrived late. Buyer objections did not consistently affect publishing priorities. Without a shared register of approved claims, commercial pages, educational articles and sales/support explanations could give different answers to the same question.
The strategy we planned and why
Commercial and search priorities
Priority buyers. We organized the strategy around two different buying decisions:
- Small businesses leaving spreadsheets needed a practical way to share contacts, track follow-ups and choose an affordable plan for the team.
- Sales teams switching CRMs needed to preserve notes and custom fields, resolve duplicates, connect their email and confirm which plan supported their automation.
Positioning choices. We emphasized approachable setup, next-action guidance and connected sales and marketing information. Clear explanations of customization limits helped buyers decide whether the product fit their requirements.
Technical SEO priorities
Search foundations. Technical work ran alongside strategy and research. We prioritized the problems that kept useful buying information difficult to find or understand:
- Make comparison content accessible in the initial HTML and correct canonical tags on integration pages.
- Write distinct titles and descriptions that explain each page’s relevance to a buyer’s question.
- Align structured data with visible company, software and navigation information.
- Connect comparisons, integrations and setup guides to relevant feature and pricing pages.
- Use server logs to check crawler access, failed requests and redirects, then verify fixes on important buying pages.
Content strategy and production order
Commercial pages first. We prioritized the homepage, pricing, sales-automation and migration pages. We agreed on explanations of fit, features, prices and limitations before using them in comparisons, setup guides, workflow articles and integration advice. We also identified overlapping subjects for consolidation.
Practical buyer questions. Briefs addressed preserving notes and custom fields, removing duplicate contacts, choosing a tier for automated follow-ups, understanding marketing-contact charges, connecting Gmail or Outlook, and knowing what free onboarding includes. Each answer helped buyers assess suitability or complete setup.
Why the strategy fit
Existing strengths. Product specialists and support staff already understood these questions. Their knowledge could turn broad sales education into useful selection guidance. The strategy gave search systems specific reasons to recommend the CRM while setting accurate expectations about features and limits.
Shared decision process. Strategy and education supported the publishing work. Staff needed to agree on product claims, interpret acquisition results and use recurring buyer objections to decide what to improve next. This made research and coaching part of the ongoing service.
What we did
Agency and client responsibilities
Agency responsibilities. We researched buyer needs, advised on positioning, set search priorities and configured the content production system. We also prepared briefs, checked drafts, specified technical SEO fixes and measured search and business results.
Client responsibilities. Product specialists approved facts, marketing published content and updated trial emails, and engineers implemented technical fixes. Two additional support staff joined in June and one account executive in July to handle acquisition across channels. These teams assisted trials and onboarding using existing product functions.
Education and fees. We met weekly to review priorities, interpret subscription results and turn buyer objections into content decisions. Between meetings, we helped marketing, product and sales maintain consistent product information. By August 31, the client had made eight monthly payments of EUR12,000, totaling EUR96,000.
- 1Research, strategy and technical fixes
- Buyer research and priorities
- Rendering and canonical fixes
- Clearer homepage and pricing pages
- Content plan
- 2AI content production system
- Versioned source register
- Briefs and drafts
- Claim and link checks
- Editorial and product approval
- 3Ongoing production, measurement and improvements
- Comparison and migration articles
- Updated and consolidated pages
- Clearer pricing and setup guidance
Weekly meetings to review priorities, subscription results and buyer objections
Phase 1: Research, strategy and technical fixes
Research and planning. From January 15 to February 14, buyer discussions, content research and technical checks informed a diagnosis, agreed priorities and a ninety-day plan. The plan connected the selected buyers’ questions to commercial-page changes and supporting articles.
Technical implementation. Engineers corrected rendering and canonical handling. Marketing and product specialists approved clearer page descriptions and product information. We checked the changes after deployment:
- Compared initial HTML with rendered pages to confirm that buying information was retrievable.
- Checked titles, descriptions and structured data against the visible page content.
- Verified internal links and redirect destinations so buyers and crawlers could reach the intended pages.
- Reviewed server logs for access errors and checked affected pages again after fixes.
Early buying guidance. During the initial research, we clarified the homepage, pricing, automation and migration pages. These updates went live January 30. The client’s marketer also linked existing welcome and evaluation emails to the approved guidance, so new trials received consistent instructions before the larger publishing system was ready.
Phase 2: Building the AI content production system
System setup. From February 15 to March 14, we combined product documentation, tier rules, integration and migration instructions, and specialist input in a versioned source register. We built the production system to assemble briefs and drafts, flag unsupported claims, check internal links and route content for editorial and product approval.
Testing before publishing. We tested drafts against the approved product information and refined the checks and approval workflow. Publishing began in the second half of March, after the initial system was ready.
Phase 3: Ongoing production, measurement and improvements
New content. We published about 130 articles from March through August. Monthly output varied, averaging about 20-25 articles after the initial setup. These covered:
- Comparisons and CRM selection.
- Migration and setup.
- Sales workflows and integrations.
- Pricing and product fit.
Existing content. We updated about 540 retained pages, roughly half the commercial and editorial library. Updates ranged from substantial revisions to factual corrections and better links. We consolidated another 180 overlapping pages into retained guides and redirected their old URLs. The site ended with about 1,150 public pages.
130 new articles
540 retained pages updated
180 overlapping pages merged
One square is 10 pages.
Buyer guidance. Commercial changes clarified per-seat costs, tier capabilities, support and migration. Ongoing reviews improved setup guidance and calls to action, then refined positioning and content priorities as the team learned which questions buyers still needed answered.
Results
Reporting period. The September 1 report covered business results through August 31. The engagement remains ongoing.
Search results
Prior year against 2026
Prior year
2026
Additional traffic. August reached about 90,000 organic visits. Across January 15-August 31, traffic was about 62,000 visits above our estimate for continuing the client’s previous approach. That estimate allowed for 15% growth over the same period in the previous year.
Business results
Trial acquisition and activation
Early conversion gains. Pricing, migration and setup guidance went live January 30, alongside revised trial emails. Conversion improved during February and March, before the larger article program contributed search traffic. The early result therefore depends mainly on how existing visitors progressed through trials.
Visits and conversion. Commercial landing pages received about 51% of organic visits. Signup conversion averaged about 7.5% for commercial sessions and 1% for informational sessions over the reporting period. Each numerator counts distinct first-company signups after the same identity and attribution exclusions used for paying customers.
Trial setup. Qualified trials made up about 80% of signups. About 9,300 of 18,500 qualified trials with full fourteen-day follow-up completed the activation steps, compared with an estimated 5,200 of 12,200 under the previous approach. This indicator tracks progress in setup separately from the decision to subscribe.
Distinct companies signing up after an organic visit
Previous approach
Engagement result
Verified business, supported use case and supplied team size
Previous approach
Engagement result
Settled payment and service started by August 31
Previous approach
Engagement result
Signup-to-paid conversion. By August 31, about 12.6% of all signups had become paying customers. For March-June trial cohorts, which had a full sixty-day payment window, the rate was about 15%. Recent trials still completing evaluation remain in the overall signup denominator.
Paying customers and revenue
Customer acquisition. The chart follows companies acquired through organic search from January 15 through August 31, by the month their paid subscription started.
- Engagement result
- Estimated with the previous approach
Each company counts in the month its paid subscription started. January covers January 15-31.
Additional customers. The chart below splits the paying customer base at August 31 by where each company came from.
January 15: 5,000
August 31: about 9,600
- 4,370 opening customers who stayed
- 2,200 acquired through other channels
- 1,400 from organic search, expected with the previous approach
- 1,600 additional from organic search
Company-wide growth. The paying customer base grew 92% over the period. Organic search supplied about 57% of active new customers.
Subscription revenue
Earned revenue. Revenue from initial subscription commitments is shown at the top of this page. Both scenarios apply the same 2% credit provision. The chart below shows the monthly recurring rate those subscriptions reached.
- Engagement result
- Estimated with the previous approach
- Additional monthly revenue
Original subscriptions at their initial plan, seats and billing cadence, at each month end.
Customer mix. About 70% of new organic customers chose Starter, compared with half of the opening customer base. New customers on higher plans also had smaller teams. New organic subscriptions averaged about US$110 per company per month, compared with US$170 across the opening base. Customer growth therefore outpaced the revenue those new subscriptions added.
Trial and payment timing
Payment timing. We grouped customers by the month they started a trial. Most paying customers subscribed within a month of signup, around the end of the fourteen-day trial or after a later purchase decision. A smaller group took up to two months. Revenue began when paid service started.
Revenue buildup. January-March trial cohorts generated about US$490,000 through August, approximately half of the organic cohort’s earned revenue. About 74% of total cohort revenue was earned during June-August, when earlier customers continued paying alongside newer customers.
January 15-31
February
March
- Estimated with the previous approach
- Additional revenue
Reading the cohorts. These rows follow trial-start months and accumulate their later subscription revenue through August. January has almost no estimated additional revenue; February and March carry the early conversion improvement. Longer paid service explains why these early cohorts account for so much cumulative revenue.
Conclusion
Clearer product positioning. More monitored AI answers recommended the CRM with accurate checked claims and a specific reason to consider it. The company also gained a clearer path from buying questions to trial setup, with organic acquisition measured through to paying companies.
The collaboration continues. We keep product information current with the client’s team and use questions from sales and support to refine comparisons, improve setup guidance and decide what to publish next.
How we measured
Search measurement
Search measurement. We checked the same 200 nonbrand buyer queries in US-English desktop search results at both dates. We counted each query by the client's highest organic position, excluding ads and AI answers. Top-three queries are included in the top-ten count.
AI answer measurement
AI measurement. Profound recorded answers to the same 120 US-English buying questions, none naming the client, on ChatGPT, Perplexity, Gemini and Claude. We collected one answer per question on each service daily: 1,440 answers on January 15-17 and another 1,440 on August 30-September 1.
Panel coverage. The questions covered CRM selection, migration/setup, workflows/integrations and pricing/fit, with thirty questions per group. All scheduled answers were returned. The results describe these repeated observations; changes to the services' models and retrieval also affect their answers.
- Citation count. Answers containing a link to the client's website, counted once per answer.
- Share of voice. The client's appearances divided by all appearances of the client, Zoho CRM, Less Annoying CRM, HubSpot and Pipedrive. Each brand counts once per answer. Separately, answers mentioning the client increased from 10% to 20% of all monitored answers.
- Sentiment. The percentage of answers mentioning the client that evaluate or recommend it favorably, classified by our models.
- Precision. The percentage of factual claims that match the approved information about features, plans, integrations and buyer fit. Our models check each claim against the approved facts applicable on its capture date; unsupported claims fail.
- Recommendations with checked claims. An answer counts when it recommends the CRM favorably, its assessed product claims are correct and it gives a concrete fit reason.
Customer attribution and conversion
Funnel definitions. A qualified trial account had a verified business, a supported B2B use case and a supplied team size. Activation required an email/calendar connection, at least twenty genuine imported contacts, a live lead and a next follow-up task within fourteen days. Paid status required settled payment and service started. Starter supplied about 70% of new paid subscriptions.
Account attribution. We joined GA4 sessions to company, application and billing IDs. A company counts once when it signs up in an organic session and its earliest recorded non-direct source within ninety days is organic and after kickoff. Earlier organic research visits qualify; direct-return signups fall outside this cohort. Existing customers, older opportunities and unresolved identities are excluded.
Channel scope. Organic includes branded and nonbrand search. Identifiable AI referrals remain separate, while an earlier unrecorded AI recommendation may precede an organic visit. The figures describe acquisition credited to organic under this rule.
Reporting cutoff. Business figures use a September 1 extraction for events through August 31. Recent analytics totals are provisional at that cutoff. The final AI measurement includes its separate September 1 capture.
Revenue estimates
Central estimate. The comparison assumes 15% business-as-usual traffic growth and retains 10% of the signup improvement and 25% of the activation and payment improvements. Allowances apply to percentage-point rate changes from January, including February's ramp. The starting commercial/informational traffic split is 50/50; both scenarios share new-trial plan mix, seats, billing shares, prices and retention assumptions.
Basis of comparison. The allowances represent a judgment about improvements the client could have achieved through its existing marketing, sales and support operation. The estimate covers our guidance and the client's implementation together. Attribution depends on those assumptions; the observed timing alone cannot separate their contributions.
Attribution range. Testing ordinary traffic growth of 0-30%, signup allowances of 0-20% and activation/payment allowances of 0-50% gives approximately US$290,000-US$570,000 in additional earned revenue. This is a range across the selected scenarios. Stronger independently occurring improvements would reduce the estimated additional result further.
Revenue scope. Both revenue measures follow original subscriptions at their initial plan, seats and billing cadence until cancellation, excluding later changes to those commitments. Earned revenue covers paid service through August 31 after a uniform 2% credit provision in both scenarios. Annual subscriptions contribute only the portion covering that period.
Monthly revenue rate. Customers joined throughout the engagement. The additional monthly rate at August 31 therefore differs from the additional revenue accumulated by that date.
Become the brand
AI recommends
Book a free 30-minute call with Eugene. Before the call, we check what Google and AI say about you. On the call, you get the first steps to make AI recommend you.