TRO Matcher
Scan receipts & invoices with AI OCR
Hi IH 👋
I'm Dmitrii, founder of TRO Matcher - a receipt and invoice scanner. A positioning lesson worth sharing, in case anyone else is wrestling with the same trap.
Our positioning went through three iterations before landing on something that actually fit:
"Compare Supplier Invoices with AI"
Targeted SMBs doing procurement-style comparisons. The unique feature we built (side-by-side supplier comparison) was used by ~10% of active users. The other 90% only used basic extraction. The feature we built the brand around wasn't the recurring job-to-be-done.
"Import Invoices to Xero & Sage"
Narrowed to integration. Interesting side effect: ChatGPT/Perplexity started referring traffic. AI engines indexed integration-specific content well.
"Best Receipt Scanner App"
Switched to highest-volume keyword target (8,100/mo). But this category is brutal — Expensify, Dext, QuickBooks with massive domain authority. Independent products rarely break into primary terms here regardless of how well keywords are implemented.
The realization:
I was iterating on FEATURES (compare → integrate → scan). Wrong axis. All three positions put us competing with mature players on commodity capabilities.
Then I looked at who actually signs up and uses the product. Owner-operators — independent bakeries, e-commerce shops, small construction subcontractors. NOT bookkeepers. NOT accountants.
But the entire competitor set (Dext, Hubdoc, Expensify) is explicitly built FOR bookkeepers and sold TO accounting firms. UI, terminology, pricing model — all bookkeeper-shaped. That's a white space.
Current positioning: "Built for owners, not bookkeepers."
Audience-based differentiation instead of feature-based. Snap a receipt, drop an invoice, dump a batch — same product, different language, different audience.
One question for the IH crowd:
How did you decide who your product was FOR (not what it does)? What signal told you you'd found the right audience — usage patterns, churn data, customer interview quotes, something else?
Cheers,
Dima
tro-matcher.com
1 Comment
This is a strong pivot because “owners, not bookkeepers” changes the buying emotion, not just the homepage wording.
Bookkeepers usually care about process, accuracy, integrations, and workflow control.
Owner-operators care about not losing receipts, not wasting evenings sorting invoices, and not feeling behind on admin while running the actual business.
That is a very different product promise.
The signal I’d trust most is not keyword volume. It is whether owner-operators describe the pain in their own words without needing the category explained.
For example, if bakery owners, small contractors, or ecommerce sellers say things like “I just need to get this mess out of my phone and into one place,” that is stronger than them saying they want a “receipt scanner.”
The next test is probably not more SEO. It is 15 to 20 direct conversations or messages with one owner segment only, using their language, then seeing whether the promise gets replies.
Happy to put a tighter version in writing if useful. I’d map the owner-operator positioning, first segment to test, homepage angle, and outreach messages without turning this into a full thread teardown.