Part 1 · Weeks 1 to 8
Positioning
Find the message worth owning.
It reads your market, groups every claim by meaning and finds the one theme only you can own. An expert locks it in consultation with you.
Platform
The system has two parts. Positioning finds the one message your brand should own. Memory checks whether buyers and AI assistants remember it. Senior experts sign off every step.
01 Read everything
The AI analyzes relevant data and information from you, your customers and competitors.
Two parts
Part 1 · Weeks 1 to 8
Find the message worth owning.
It reads your market, groups every claim by meaning and finds the one theme only you can own. An expert locks it in consultation with you.
Part 2 · Every week after
Check that the market remembers it.
Four signals add up to one Memory Score out of 100, shown next to your cost to win a customer (CAC). It's reviewed weekly, with your next moves ranked.
Explore MemoryThe locked message is what Memory measures against.
Why does it matter
Your leaders, marketers and sales team tell the same story on every channel.
Buyers ask AI first now. We check ChatGPT, Claude, Gemini and other AI tools so they describe you clearly.
Memory Score shows whether buyers remember you, and whether new customers are getting easier to win.
Tech stack
A chatbot answers what you paste into it. Ours was built for one job: accurately finding and tracking the brand position of B2B tech companies.
Methodology
Every brand story is built from six pieces. If you don't get all of them right, you will struggle to create a unique position for yourself in the market.
The problem you solve for customers that no one else solves as well.
The market where your strengths stand out, in words AI tools can repeat.
The product or service your customers would opt for if you didn't exist.
What your strengths do for customers, in their words, backed by evidence.
Your happiest customers, and what they like most about your brand.
The one feeling buyers should have when they think of you.
The single claim only you can make.
Guardrails
Every number is calculated in code. Models sort and write, nothing more.
Every number in a claim has to exist in the evidence, or it gets flagged.
Put a competitor's name in your story. If it's still true, the AI won't recommend it.
One loud review isn't a trend. A need counts only when enough customers say it.
Your own customer interviews count most, then review sites, then Reddit and search.
If your leadership rules something out, the message won't go there.
GTM workflows
Our recommendation engine reads your scores and suggests what to work on next every week.
Every week it recommends what to change in:
Where the narrative should run, and how to set up each channel for it.
What to write or rewrite so every piece carries the narrative.
Fixes to your site so search engines and AI assistants can read it clearly.
Case study
Where SubscriptionPro stood when we started, and the one-message plan their team is now putting to work.

Trust and privacy
Intake answers, documents, customer quotes, chats and account connections.
Emails, phone numbers and card numbers are stripped out before any text reaches our AI.
Team members see only the brands they're given access to.
File downloads included. An admin can review the log.
Optional two-step sign-in. No public sign-up is possible.
We can delete your uploads and transcripts at the end. Your scores and themes stay.
No. A-square is a specialized service, not a tool. It's different in four ways. 1. With a chat tool, you're left to judge its answer on your own. With us, experts who know your field work on every engagement. They interview your team, check what the AI finds and define the final answer with you, so what you get is accurate. 2. A chat tool reads a long pile of content the way people skim: closely at the start and the end, less in the middle. Parts of your content never make it into the answer. Our system uses statistical methods that treat every post, page and review as an equal data point, so every theme gets counted, wherever it sits in the pile. 3. A chat tool gives you one model's answer with no test behind it. The only check is whether it sounds right to you. We tested many AI models on cases where the right answer was already known, and the best one runs each step of the analysis. When new models come out, we test again and update our method and guardrails. 4. A chat model is a black box. Ask for the evidence behind an answer and there's little to point to. Every result in our system keeps its trail: what content was read, which method ran and why the conclusion came out that way. You can show it to your team as it is. Because experts lead the work and every step is tested and recorded, you start with a system that's ready to speed up your results.
Positioning runs first and ends in one recommendation, which an expert locks in consultation with you. That message becomes the baseline Memory measures against each week. Memory then ranks the next actions, so the two keep feeding each other.
One brand story that only you can own. Positioning is designed to reach it from three directions: your firm (what you publish, what your teams say and where you're strongest), your competitors (up to eight, to see what's already claimed) and your customers (interviews, reviews and search, to see what they actually value). The result is one recommendation card that brings it all together: who the story is for, the pain it solves, the category you compete in, the value only you can claim, the feeling buyers should have, and the proof behind it. Every line links to its evidence.
No. The AI can draft a fix, but only an expert can apply it, and the message itself is locked in consultation with you.
Positioning usually takes 1 to 8 weeks. How fast depends on getting the right data from your team on time, so we can gather, standardize and upload it without delays. Over those weeks we also interview your teams, run the analysis and lock the narrative with you, which sets your Memory Score and CAC baseline. From week 9, an expert reads your Memory Score every week, and your brand signal is refreshed quarterly to keep pace with the latest AI models and market trends.
Five kinds of source. Your connected tools are read again on every refresh. First-party insight: transcripts and notes from our expert interviews with your product, marketing, sales and leadership teams and your external partners, plus customer interviews, sales call transcripts and surveys. Your brand: your website, social posts, ads and emails, through read-only links to HubSpot, Google Analytics and Search Console. Competitors: up to eight, with their sites, posts, ads and G2 profiles, which we collect ourselves. Customers: reviews on G2 and Capterra, Reddit and search queries. AI assistants: ChatGPT, Claude, Gemini and other AI tools, asked once with web search off and once with it on.
At every step, and they bring what AI can't: judgment, experience and deep expertise in your market. First, they interview your teams and prepare your data. The AI then reads every source and shows every analysis it ran, so nothing is a black box. Experts stress-test those findings against their own experience. When something looks off, they adjust the AI's settings and re-run it, so it answers the exact questions they want answered, fast. Next, they consult with you to work through the fine points, and only then lock the signal. After that, the AI scores the four signals every week and an expert decides what to change.
The more you share, the sharper the insight we can find for you. Ideally, that means links to your website and social accounts, and read-only access to Google Analytics, Search Console and your CRM. Add decks and documents, customer interview transcripts and surveys, sales call transcripts, and time with your product, marketing, sales and leadership teams for expert interviews. We collect competitor, review, Reddit, search and AI-assistant data ourselves. Missing some of it is fine: we start with what you have, and every extra source sharpens the picture.