What our leads are actually telling us

Apr–May 2026 lead data from Supabase, my read on what we should keep doing, what we should drop, and how we should think about the rebrand.

The 60-second read

If I strip out tests, vendor pitches, white-label outsourcing shops, and job seekers, we got 16 real buyer leads in seven weeks. That's the real number. Anything I quote from a raw export is misleading because 28 of 55 submissions were people trying to sell us something or asking for a job, not asking to buy.

Three patterns jump out and they should drive every decision we make in the next 90 days:

1. Ecommerce listings / product ops is our biggest gravity well. 6 of 16 real buyers (38%) came in for some form of "add or clean up products on Shopify/eBay/Amazon/Lightspeed." Crate & Barrel is in this bucket and is the largest opportunity in the dataset.
2. Data cleansing / extraction is the second pull. 4 of 16 (25%) — OCR, dedupe, CRM scrub, HubSpot data entry. Tight scope, clear deliverable, easy to close.
3. Accounting / bookkeeping is the third. 2 of 16 (13%) — recurring monthly work. Smaller deals but higher LTV.

That's 76% of real demand coming through three verticals. Everything else is one-off (one VA request, one HR exit-interview request from JK Lakshmi, one staffing scale-up from Revguard). The website currently advertises 50+ services across 22 industries. The mismatch between our surface and our demand is the rebrand story.

The one thing I want Chakshu to take away

What's in this document

• 1. This summarypage 1
• 2. April leads, every one of thempage 2
• 3. May leads (through the 19th)page 3
• 4. What the 16 real buyers actually wantpage 4
• 5. Where they're finding us (and what's missing)page 5
• 6. Pages and services I'd act onpage 6
• 7. The rebrand, framed as a positioning decisionpage 7
• 8. The 90-day plan to triple real-buyer leadspage 8

April: 39 submissions, 11 real buyers

28% real-buyer rate. Most of the noise was vendor cold pitches and white-label outsourcers trying to subcontract from us.

April real buyers

April noise

May 1–19: 16 submissions, 5 real buyers

Smaller volume than April so far, but real-buyer share is 31% (up from 28%). And the May leads punch above their weight.

May real buyers

May noise

What the 16 real buyers actually want

When I group every real buyer ask by what they actually need delivered, three categories cover 76% of demand.

1. Ecom listings & product ops · 38% of real demand

This is the gravity well. Six buyers across two months, asking for variations of the same thing: "please load, format, and clean up products in our store." The platforms vary (Shopify, eBay, Amazon, Lightspeed, Magento), the volume varies (5 listings to "thousands of PDPs"), but the work is the same — humans-in-the-loop catalog operations.

• Justice Rines: curate & add products
• Steve Thomas: Shopify CSV imports
• Shawn Coughlin: eBay listings paid test
• Dave Knudtson: Lightspeed → Shopify migration
• Philip Litt: 500 watches with photos + descriptions
• Sabrina Schaefer (Crate & Barrel): PDP copy at thousands-of-products scale

What I think: we should own this category publicly. We do the work, we have case studies (Geoff Hopkins / National Workwear's Shopify migration is on our site), and the demand is sitting there. Today our nav buries it under "E-commerce Outsourcing" with seven sub-services. I'd promote it to the top of the homepage.

2. Data cleansing & extraction · 25%

Tight scope, clear deliverable, easy to quote. Cameron Gold's 10K-image OCR is the headline — concrete unit count, concrete output format, explicit pricing question. Tim Onoff's 2,000-row consultant DB scrub is the same shape. Robert Tuffin and Courtney Yeager want recurring HubSpot/CRM data entry which is also this category.

What I think: we have service pages for data cleansing, data extraction, data conversion, data verification — five different pages on the same theme. We should consolidate them into one strong page that ranks well, with sub-sections for the specific deliverable types.

3. Accounting & bookkeeping · 13%

Two buyers, both recurring monthly work — Big Burro's real-estate bookkeeping and Michael Williams' CPA tax filing. Smaller deal sizes but higher retention.

What I think: our accounting pages should foreground the recurring/monthly model. Today they read like project pricing. The buyers in this bucket want a partner, not a vendor.

The outliers (4 leads, 1 each)

Logan Gross/Revguard (scaling 100–200 FTE — above ICP), Shikhar Johari/JK Lakshmi (Indian enterprise exit interviews — wrong geography), Jack Lane/Oak & Key (real-estate VA, 5 hrs/wk — too small), Jeremy Adan/BuildSiteMarketing (cold callers — a different service line).

What I think: outliers are not problems, they're noise that confirms the core. We don't optimize the site for any of them. If the rebrand narrows our nav to the three core verticals, these still find us through long-tail SEO pages and we close the ones that fit. Revguard might actually be a strong fit; we should follow up if we haven't.

How I classified each lead (for audit)

A submission counts as a real buyer only if all three are true:

1. They have a concrete service request (a thing they need done).
2. They are asking us to do work, not offering to do work for us.
3. They are not asking for a job, not a freelancer selling themselves, not a vendor selling us a tool.

White-label outsourcers — Indian or Pakistani BPO shops saying "we are looking to outsource and collaborate" — count as junk. They look like buyers in raw exports; they're not. They want our work, not our service.

Where they're really finding us

Looking only at the 16 real buyers — not the noise — here's what referrers tell us about our discovery surface.

What this means for our page strategy

Two things, both backed by data above:

(a) Google organic still dominates volume (44%) but the queries that bring real buyers are commercial-intent, not informational. Justice came in for "product curation," Robert for "HubSpot data entry assistant UK," Courtney for "donor pledge data entry." These are tight buying queries. The right surface for them is our service and integration pages — not blog posts.

(b) The four "Direct, no referrer" buyers are not actually direct. They came from somewhere we can't see — LinkedIn mobile in-app browser, Slack shares, bookmarks from prior visits, podcast bios. With UTM-tagged links (which we now have the tooling for — task #17 shipped May 15) we'd actually know.

What I know we're missing

• No share buttons on blog posts. If our content is good enough that Gemini cites it, we should at least let humans share it. A LinkedIn share button alone could materially change the inbound mix.
• No UTM discipline on Chakshu's LinkedIn profile. His "Website" field on LinkedIn points to acelerartech.com with no UTM. Every click from his profile counts as "Direct." We're undercounting LinkedIn.
• No email signature link standard. Same problem. Every reply from anyone on the team that ends with "— Chakshu | acelerartech.com" generates "Direct" traffic on click. Add UTMs and we'd see it.
• No "Why X teams choose Acelerar" page for the three real verticals. Ecom listings, data cleansing, accounting — each deserves a focused destination, not just one of 50 navigation items.

None of these are big projects. The UTM tooling is already built. Adding share buttons is half a day. Updating LinkedIn bio fields is 15 minutes. The three new positioning pages are a 2-week content sprint with the writer skill we already have.

Pages and services I'd act on

Grounded in the lead data above — not in what's trendy in BPO marketing.

Pages we already have that are pulling — keep and reinforce

Based on the real-buyer asks, these existing pages are clearly doing work even if we can't measure conversion attribution per-page yet (we will, after task #18 deployed today):

• /ecommerce/amazon-listing-services/ · /ecommerce/shopify-product-upload-services/ · /ecommerce/ebay-listing-services/ — ecom buyers keep finding us. These pages work.
• /accounting/small-business-bookkeeping-services/ — Big Burro's ask landed here.
• /data-entry/data-processing-services/ · /data-management/data-cleansing-services/ — Tim Onoff, Robert Tuffin's pattern.
• /blog/top-data-entry-outsourcing-companies-2026/ — but with a caveat: this page generated a job-seeker lead (Rajeswari) which means it's attracting "data entry jobs" searches alongside "data entry providers." We may want a stronger qualifier in the H1 ("for buyers" not "for jobs").

Pages I'd create now — 3 priority builds

Pages I'd quietly de-emphasize

Not delete — keeping for SEO long tail — but pull from main nav:

• The 12 "industries" pages, except Retail/Ecom, Real Estate, and Financial Services (the only three that match real-buyer geographies).
• "Recruitment Process Outsourcing" — no real buyers came in for this. Revguard's "scale to 100–200 employees" is the closest, but he wasn't asking us to recruit.
• "AEO Services" and "SEO Services" — we're a BPO selling SEO services on the same site. It confuses positioning.
• "AI-Native Outsourcing" as a top-line concept — keep the page, but stop making it a nav item. None of our real buyers found us via that framing.

Services I'd add to the menu (because buyers asked)

• Exit-interview administration (Shikhar/JK Lakshmi). Small market in India specifically. Worth a single page; not worth a top-nav slot.
• Cold-call outsourcing for marketing agencies (Jeremy/BuildSiteMarketing). We don't currently advertise this. Real demand exists.
• Donor & nonprofit data operations (Courtney/United Way). This is a quietly large vertical — nonprofits have data work, no internal capacity. Worth one page.

None of these three replace the core three. They sit on the long-tail page list for SEO.

The rebrand, framed as a positioning decision

I want to disagree with the framing first, then propose a version of it I think we should do.

What I think a rebrand should NOT be

• A new logo. The current mark is fine. Changing colors costs us SEO and recognition with no commensurate gain.
• A new company name. We have brand equity, ISO certification, case studies, and a sitemap built around "Acelerar Technologies." Renaming triggers months of redirect work and admin overhead.
• A "premium repositioning" that abandons our small-business buyers. Most of our 16 real buyers are mid-market and small (median deal under $5K/mo). A premium-only stance kills 80% of inbound.

What I think a rebrand SHOULD be

A focus exercise, not a redesign. The data above says we have demand in three places — ecom ops, data ops, finance ops — and we're trying to be visible in 50. The rebrand is choosing what the homepage emphasizes, what the nav shows, and what we put on our LinkedIn header.

That's the elevator. Everything else we offer stays on the site, ranks for long-tail queries, and gets surfaced when the buyer's specific search finds it. But the homepage, the LinkedIn header, the sales deck, and every paid ad lead with the three categories.

What changes in practice

1. Homepage hero copy. Today it lists everything. New version: "Three back-office disciplines we obsess over: ecom catalog ops, finance ops, data ops." Plus the three category tiles with case study proof for each.
2. Main nav. From "Services / Industries / Resources / Pricing / Contact" to "Ecom Ops / Finance Ops / Data Ops / Pricing / Resources / Contact." Industries dropdown collapses into a Resources sub-menu.
3. Footer. Keep all 50 service pages linked from the footer for SEO. They're not gone, they're just not in the main nav.
4. LinkedIn header image & profile. Same three-category framing.
5. Sales deck cover. Same.
6. "How we describe ourselves" snippet. Update the meta description across all pages so it reflects the new positioning. SEO impact: positive — narrower keyword set, higher relevance.

What we keep — the brand assets that already work

• The Acelerar name and word mark.
• The royal-blue + coral palette (it's our existing brand and ranks well in user testing for B2B trust).
• ISO 27001 + SOC 2 trust signals on every page.
• The 7-day deploy / 99.5% accuracy / $6–8/hr pricing claims — these are core differentiators per Chakshu's framing and we've proven them.
• The case studies (Eric/Elite Commerce, Carol/On Air Dining, Geoff/National Workwear, Ryan/Brafton, Jonathan/Zynga, Jesus/Scopely). They cover all three of our chosen verticals.

The single sentence I'd put on the rebrand brief

How I'd triple real-buyer leads in 90 days

From 8/month real buyers (our Apr–May run rate) to 24/month. Specific actions, ordered by effort vs return.

Weeks 1–2 — Fix attribution and discovery infrastructure

Most of this is already done as of today (May 19). What's left:

• UTM-tag Chakshu's LinkedIn profile, X bio, and email signature. 30 minutes total. The tooling shipped May 17. We just need to use it.
• Add LinkedIn share buttons to the top 10 blog posts and all three new positioning pages. Half a day of work. Doubles the chance of any single post going LinkedIn-viral.
• Verify the new attribution Slack ping is working for every new lead. (It is — confirmed today.)
• Add a "How did you find us?" follow-up to every sales call — Chakshu & team write it down. Compare against form data to see how often the dropdown lies.

Weeks 3–6 — Ship the three positioning pages

• /ecommerce/catalog-operations-at-scale/ — 1 week, includes National Workwear case study + Crate & Barrel-shape ask in the FAQ.
• /data-management/document-data-extraction/ — 1 week, mention HEIC/JPG/PDF, include the Gold Wing 10K-image example as a use case.
• /accounting/property-management-bookkeeping/ — 1 week, vertical-specific with Big Burro 170-unit shape as the headline use case.
• Refresh the homepage to lead with the three categories. 1 week including content + dev.

All four use the acelerar-writer + acelerar-editor skills we already have. Expected: 4 weeks total, sequentially.

Weeks 7–10 — BOFU content sprint

Per Sam Dunning's playbook (the one Chakshu approved earlier). Ship 5 BOFU pieces aimed at the queries our real buyers used:

• "Shopify product upload outsourcing: 5-listing test approach" (ShoreLine pattern)
• "How to outsource 1,000+ PDPs without losing brand voice" (Crate & Barrel pattern)
• "Bulk image-to-CSV: how OCR-plus-human-review works at scale" (Gold Wing pattern)
• "Monthly bookkeeping for property managers: what to expect" (Big Burro pattern)
• "HubSpot data entry assistants: when to hire, when to outsource" (Robert Tuffin pattern)

Each is a 2,500-word piece optimized for the AI-citation patterns that worked for Crate & Barrel. Each can ship in 2 days through our content pipeline.

Weeks 11–12 — Measure and iterate

• Compare Jul–Aug 2026 real-buyer leads to Apr–May baseline.
• Look at which of the new pages is converting fastest (we now have entry-page attribution).
• Look at which UTM-tagged surfaces are driving the most real buyers (we now have proper UTM data).
• Look at which "How did you find us?" answers are most associated with real buyers vs noise.
• If LLM-attributed leads grow as a share, double down on AI-citation content. If LinkedIn UTMs dominate, double down on Chakshu's profile activity.

What success looks like by August 31

The baseline (Apr–May 2026): 8 real buyers/month, 51% junk share, 1 LLM-attributed lead/month. The targets are 3x on the first metric, half on the second, 3x on the third. None require us to add headcount, change the logo, or change the company name. They require us to choose.

What I'm uncertain about

• 16 real buyers is a small sample. Patterns could shift in June with more data. The plan is structured so the first two weeks of work make us better at measuring whether the patterns hold.
• I don't have sales-cycle data on these 16 leads. I know who came in; I don't know who closed, at what value, with what cycle length. Chakshu has that in his head. Comparing actual closed deals to my "real buyer" classification would tighten this analysis materially.
• The Crate & Barrel inbound is one data point. It could be the start of a trend or a one-off. The way to find out is the BOFU content sprint above — if we ship five more LLM-citation-optimized pieces and three more enterprise-shape leads arrive, the pattern is real.
• If Chakshu still wants a logo refresh, that's a separate conversation. I'm not opposed; I just don't think the data supports prioritizing it over the positioning work above.