Portfolio case study · Updated August 24, 2026
ShopProp Realty is a luxury flat-fee real estate brokerage, managing-broker-led since 2007, operating across 8 states. It is also the oldest and largest proving ground for TruePrime Go — the first business where every feature was built, tested, and refined before being offered to outside customers.
This is not a polished marketing case study with inflated numbers. It is a transparent look at what Go actually does for a real business, every day — including what worked, what failed, and what we changed as a result.
| Detail | ShopProp Realty |
|---|---|
| Founded | 2007 |
| Type | Luxury flat-fee real estate brokerage |
| States | 8 (and expanding) |
| Model | Managing-broker-led, flat-fee commissions |
| Challenge | Cover hundreds of local markets with relevant content, respond to inquiries faster than competitors, track market-level performance — without a large marketing team |
| Go program since | Day one — ShopProp was the original build target |
Real estate is a local, high-intent business. Buyers search for specific cities, neighborhoods, and property types. Sellers compare brokerage fees and commissions. The challenge: covering hundreds of local markets with relevant content, responding to inquiries faster than competing brokerages, and tracking which markets are growing versus declining — all without a large marketing team.
Before Go, ShopProp faced the same problem most service businesses face: the work that generates leads (content creation, SEO, review management) competes for time with the work that closes deals (showing homes, negotiating contracts, managing transactions). Marketing always loses that fight because it does not have a deadline — tomorrow's lead feels less urgent than today's closing.
The result: inconsistent content, slow responses to website inquiries, and no systematic visibility in the markets where ShopProp operates. The business grew through referrals and reputation, but the online discovery layer was thin.
Go generates and maintains landing pages for cities and markets across ShopProp's 8-state footprint. Each page targets local search queries — "flat-fee real estate in [city]," "luxury listing agent [city]" — with content specific to that market.
These are not thin templates with a city name swapped in. Each page includes locally relevant information: market context, service details specific to that state's real estate laws, and content depth that makes the page genuinely useful to someone researching that market. Go's enrichment cycles continuously improve pages — adding depth, updating market context, and strengthening internal linking based on ranking performance data.
The volume is significant: hundreds of pages. But the content standards are strict. Every page must pass a quality test: "Would a homeowner actually learn something from this?" Pages that fail get enriched or consolidated — never left thin.
ShopProp competes against traditional brokerages, flat-fee competitors, and low-cost listing services. Go builds honest comparison pages for each — what ShopProp does better, where the competitor has genuine advantages, and what questions a seller should ask.
These comparison pages are the strongest AEO content type we have found. When someone asks an AI assistant "ShopProp vs. [competitor]" or "[competitor] alternative," a well-structured comparison page is exactly what gets cited. The structure — clear, honest, side-by-side — gives AI assistants the format they need to formulate accurate recommendations.
Honesty matters here. The comparisons acknowledge where competitors win. This is not just ethical — it is strategically sound. AI assistants evaluate source trustworthiness, and balanced comparisons signal higher authority than one-sided marketing pages.
Go maintains ShopProp's full AEO stack: llms.txt (a machine-readable content summary for AI assistants), brand-facts.json (structured business data — services, pricing, locations, and verifiable claims), and JSON-LD schema markup on every page.
When ChatGPT, Perplexity, or Gemini evaluates whether to recommend ShopProp for a real estate query, these artifacts provide the structured, reliable information AI engines prefer over unstructured marketing copy. The brand-facts file states exactly what ShopProp does, in which states, at what price — in a format machines can parse directly.
This AEO layer is why Go was built. Traditional SEO tools do not produce these artifacts. Most agencies have not developed AEO capabilities. ShopProp was the proving ground for every AEO technique now included in every Go customer's program. Full AEO guide →
An AI agent on ShopProp's website responds to visitor inquiries in seconds — whether someone asks about listing fees at 2 PM or property availability at 2 AM. The agent is trained on ShopProp's brand facts: pricing, service areas, commission structure, and process. It qualifies leads, captures contact information, and routes urgent inquiries to the right team member.
The response time advantage matters enormously in real estate. Research consistently shows that lead conversion drops dramatically after the first 5 minutes. An AI agent that responds in seconds captures leads that a "we'll call you back within 24 hours" workflow loses to faster competitors. In real estate specifically, the first agent to respond often wins the listing appointment.
Go tracks ShopProp's keyword rankings across hundreds of real estate search terms, segmented by state and market. Twice-weekly intelligence reports highlight: which markets gained or lost ranking positions, which keywords hit the quick win zone (positions 6–10 where a content push could reach page one), which competitor pages are rising, and what actions were taken in response.
The reports focus on outcomes and recommended actions — not just raw data. The question each report answers is: "What happened in our markets this week, and what should we do about it?"
Go scans ShopProp's review profiles daily across platforms. When new reviews appear, drafted responses are queued for human approval. Positive reviews get timely acknowledgment. Negative reviews get thoughtful, professional responses — drafted by AI, reviewed by humans before publishing. Nothing goes public without a human eye.
Review velocity matters for both SEO (Google factors review recency into local rankings) and AEO (AI assistants evaluate business reputation when making recommendations). Consistent, professional review responses signal an active, well-managed business.
Running Go on our own business taught us things we could not have learned building for abstract customers. As of late August 2026, the TruePrime Go program on trueprime.ai itself has grown to 120 live pages — guides, comparison pages, blog posts across 10 industry verticals, and AEO artifacts — all built, deployed, and verified by the same system customers use. These lessons shaped every Go customer's experience:
Early on, we generated city pages from templates — swap the city name, adjust a few details, publish. The pages existed, but they did not rank well. Worse, we discovered this approach carries compliance risk: Google classifies pages with 90% or more shared template content as "doorway pages" — a form of spam that can trigger penalties.
When we started enriching pages with genuinely unique, locally relevant content, rankings improved measurably. This became a core principle: Go's enrichment cycles continuously improve existing pages rather than just creating new ones. Every page must have at least 20% unique content after normalization. How the enrichment process works →
We learned the hard way that regenerating a page that is already ranking can cause the ranking to drop. Google re-evaluates the page, and during that re-evaluation period, positions often decline temporarily — and sometimes permanently if the new version is evaluated differently.
The stability rule — never regenerate a page ranking in the top 20 — was born from watching our own rankings temporarily drop after unnecessary content refreshes. Now, pages that rank well are protected. Enrichment happens only on underperforming pages. This rule applies to every Go customer.
At one point, we had multiple scripts tracking the same metrics with slightly different methodologies. The numbers disagreed. This is worse than having no data — it creates confusion about what is actually happening and undermines trust in every report.
Now: one measurement script per metric, one source of truth, no exceptions. If two numbers disagree, one of them is wrong, and the wrong one needs to be eliminated — not reconciled. This discipline carries through to every Go customer's reporting.
When we built ShopProp's comparison pages against competing brokerages, they consistently outperformed other content types for AI citations. The structure — clear, honest, side-by-side with strengths and weaknesses for both parties — is exactly what AI assistants need to formulate accurate recommendations.
This finding drove us to prioritize comparison pages for every Go customer. Within the first month, every customer gets comparison pages against their named competitors. See our own comparison pages →
ShopProp's Go program generates hundreds of pages. That scale matters for coverage — you cannot rank in a city if you do not have a page for it. But scale without quality creates serious risk.
We experienced this directly. An early generation cycle created thousands of keyword-variant pages with nearly identical template content. The pages existed, but they added no value. We had to carefully unwind this — adding noindex directives, consolidating thin pages, and implementing strict quality gates. The experience led directly to the content standards now applied to every Go customer. How Google evaluates new sites →
In one build cycle, a generation script created pages and added them to the sitemap — but the pages never actually deployed to the live server. Google crawled the sitemap, found the URLs, got 403 errors, and recorded those URLs as broken for months. Recovery took longer than the original build.
Now, every deploy includes mandatory verification: spot-check live URLs, confirm 200 responses, and verify the sitemap only contains pages that actually exist on the live server. A deploy without logged verification did not happen. This rule is absolute for every Go customer.
| ShopProp lesson | What it means for Go customers |
|---|---|
| Enriched pages beat templates | Go enriches your existing pages continuously — depth over volume |
| Stability rule | Pages ranking well are never regenerated — your rankings are protected |
| One source of truth | Each metric has exactly one measurement — no conflicting dashboards |
| Comparison pages for AEO | Competitor comparison pages built early for every customer |
| Quality over volume | Every page must pass a genuine usefulness test — no thin templates at scale |
| Deploy verification | Every URL verified live after every deploy — no sitemap lying |
The original six lessons above were discovered in ShopProp's first months. Since then, running Go across three portfolios has surfaced additional patterns that reinforce and extend those early findings:
We observed that AI engine visibility scores correlate directly with content freshness. When page publishing paused for several weeks, AEO scores declined measurably. When content velocity resumed — even before new pages had time to fully rank — the floor stabilized. The signal of active publishing appears to be a factor in AI engine scoring, independent of any individual page's ranking.
This is why Go maintains a steady content cadence for every customer rather than building in bursts. Consistency beats intensity.
Across the portfolio, blog posts consistently reach search indexes within days to two weeks, while dedicated landing pages take four to six weeks on some engines. This led to a two-track indexing strategy: publish a blog post first for immediate search signal in a new market, then build the corresponding landing page for long-term presence. Both formats are needed — the blog provides speed, the landing page provides permanence. Why new sites take time to index →
The original lesson about template pages taught us that quality must be enforced systematically, not case-by-case. Go now runs automated compliance checks on every deploy: pricing verification against the current facts sheet, internal link validation (every link target must return 200), banned-word scanning, and minimum unique content thresholds. A page that fails any check does not deploy. This catches problems before they multiply across hundreds of pages. Marketing audit checklist →
TruePrime Go's own marketing site is now the third portfolio proof point alongside ShopProp and AskBeforeYouEat. With 96 pages covering product positioning, 7 competitor comparisons, 10 industry verticals, an AEO artifact suite, and a blog publishing twice weekly — the site demonstrates exactly what a Go customer receives. When a prospect asks "what does this actually produce?" we point at the site they are already reading. Q3 marketing reset checklist →
TruePrime Go launched its own site on July 8, 2026. By Day 78, we had 120 live pages — 10 guides, 8 comparison pages, 98 blog posts across 10 industry verticals, and 4 supporting pages. Here is what the first eleven weeks revealed that we did not expect:
We built comprehensive content covering every tracked keyword family within the first month. By Day 30, we had more pages than most competitors. By Day 78, we still had zero Google rankings across all 34 tracked keywords — though Bing has indexed approximately 50 pages. On August 21 — Day 75 — we ran a formal technical diagnostic. The finding: no server-side blockers (no X-Robots-Tag noindex headers, no Googlebot blocks, canonicals correct, sitemap valid). The primary suspect is a missing robots.txt at the domain root — a 30-second fix requiring domain-level access that has been pending for 24 days.
The lesson is sharp: 120 pages of quality, compliance-audited content mean nothing if search engines cannot discover them. Content production is necessary but not sufficient. Discovery infrastructure — robots.txt, sitemap submission via GSC, directory citations, external links — is a separate requirement that cannot be skipped or deferred. We deferred it and paid with 78 days of invisibility.
Notably, Bing indexed our pages without intervention while Google has not. This confirms the content itself is crawlable and indexable — the gap is specifically in Google's discovery of the site, likely tied to the missing robots.txt (which provides the sitemap pointer) and the absence of Google Search Console submission.
We built automated quality gates from Day 1: pricing verification against the current facts sheet, banned-word scanning, internal link validation (every target must return 200), schema validation, and unique content thresholds. After 120 pages and dozens of build cycles, zero compliance violations have reached production. The GPACH (Go Program Automated Compliance & Hygiene) audit runs weekly and has maintained a clean record since launch — the August 21 audit returned 0 WARN across every check category. The gates prevent template bugs from multiplying across the site, which is exactly the failure mode that created the ShopProp keyword page incident.
In the first 30 days, new page creation was the highest-leverage activity — covering keyword families with no content. After Day 30, with most keyword families covered, enrichment cycles became more valuable: deepening existing pages with TCO tables, day-by-day timelines, industry-specific guidance, and competitive analysis. An 18KB page with three unique data tables is more authoritative than three 8KB pages covering the same topic from slightly different angles. Depth beats breadth once breadth is established.
By Day 78, the enrichment program runs continuously: every growth cycle refreshes the oldest comparison pages with current market context, deepens guide pages with new FAQ sections, and adds cross-links between related content. This keeps the content library fresh — a signal both search engines and AI assistants weight for recommendation decisions.
In 78 days of tracking, we observed 50+ SERP position changes among competitors for our tracked keywords. New AI-branded competitors appeared and entrenched (marketengine.ai, aeoengine.ai, agencyaeo.com, ranked.ai). Reddit threads rotated in and out of top positions — as of late August, Reddit holds #1 on 4 of our 34 tracked keywords (down from 7 two weeks ago as real competitors consolidate). The competitive landscape is volatile: on a single scan day, 7 keywords had a different #1 holder than three days earlier. Enterprise players are entering: Salesforce appeared at #3 for "ai marketing team," pushing Jasper from #3 to #5. The lesson: competitive content needs active monitoring and regular refreshing, not just initial build.
These lessons are still unfolding. At Day 78, the TruePrime Go program has grown to 120 pages — 10 guides, 8 comparison pages, 98 blog posts across 10 industry verticals, and 4 supporting pages. Every page has passed through GPACH automated compliance gates (pricing accuracy, banned-word scanning, internal link health, schema validation, unique content thresholds). The most recent weekly audit — August 21, 2026 — returned 0 WARN, 0 CRITICAL across all pages.
And yet: zero Google rankings across 34 tracked keywords. Zero AI engine citations. Seventy-eight days of publishing quality content into a void — though Bing's indexing of 50 pages confirms the content is technically sound and crawlable.
We are being transparent about this because it is the most important lesson the program has produced so far.
Go handles execution. ShopProp's humans still handle everything that requires judgment, relationships, and physical presence:
The division is clear: Go generates visibility and captures leads. Humans convert leads into relationships and transactions. Neither replaces the other.
When evaluating any marketing product, the strongest question is: "Does the company use it on themselves?"
We do not just sell Go — we depend on it. ShopProp's lead pipeline runs through Go. If the system broke, we would feel it in our own revenue before any customer would. AskBeforeYouEat (our consumer nutrition brand) runs on Go as well. The pages you are reading right now — on trueprime.ai — are our own Go program.
Three businesses, one system, same quality standards. That is a stronger guarantee than any testimonial we could write.
One of the most important patterns we have observed across the portfolio: AI marketing compounds. The content library grows, domain authority accumulates, data feedback loops sharpen keyword targeting, and AEO citation signals snowball as more pages become citable. Businesses that quit at month 3 — right when compounding starts — leave the most valuable phase on the table. The pages built in month 1 are still generating value in month 12 and beyond. The compounding effect explained →
As of Day 78, TruePrime Go's own site has zero Google rankings and zero AI engine citations — but Bing has indexed approximately 50 pages. We have 120 quality-audited pages, clean compliance records, and a valid sitemap — but a missing robots.txt at the domain root likely prevented Google from ever discovering the site. We are fixing the discovery infrastructure now.
We include this because it is real. A marketing company that hides its own struggles is selling fantasy. The content library is an asset that compounds — the moment search engines like Google discover the site, 120 pages of quality content are ready to compete. Bing's indexing validates that the content works. The lesson is not that the system failed. The lesson is that content quality alone is insufficient without discovery infrastructure: robots.txt, GSC submission, directory citations, external links. We missed the plumbing. We are fixing the plumbing.