MENA is not where this product began. It is where the research pointed.
Roya Plus started as an AI-powered brand analysis tool. Point it at a company, get a report back. The founding team had a working prototype and a reasonable assumption: business owners want to understand their brand's digital position, and AI can generate that understanding cheaply.
My job was to validate the product direction before the team committed engineering time to it. Not to decorate the assumption — to test whether it survived contact with the people who'd have to use the thing.
The competitive analysis surfaced the problem quickly, and it wasn't a feature gap. SimilarWeb, SEMrush, Ahrefs, Lucidya — every serious tool in the category is built for marketing specialists. They output data that helps a specialist decide what to do next. That's a coherent product for a competent audience.
But it means the output is only valuable to someone who already knows how to read it. A business owner opening a SimilarWeb dashboard gets numbers, not decisions. The category had no answer for the person paying for marketing but unable to evaluate it.
A full capability comparison across the major platforms, mapping what each one does, who it serves, and where the coverage stops. The consistent blind spot: Arabic-language digital intelligence, and specifically dialect-level understanding. Not "does it support Arabic" — most claim to — but whether it can tell Gulf Arabic from Egyptian, which is the difference between a usable sentiment read and a decorative one.
Four depth interviews with digital marketing practitioners working in the MENA region, each running the full arc: current tool stack, actual workflow, where deliverables come from, what they'd pay for, what they'd never trust. Participants are anonymised here as User 1 through User 4.
Four interviews is a small n, and I want to be honest about that — this was directional research meant to kill bad assumptions fast, not to size a market. It did that job. Three of the five eventual pivots came directly out of contradictions between what the team believed and what all four participants independently described.
The team had designed for business owners. Every participant described the same reality: marketing practitioners are the ones who generate these reports — as deliverables for their clients. The business owner is the audience, not the operator. Designing the interface for the audience would have made it unusable for the person actually driving it.
Social media data was described as non-negotiable by all four. In this market the brand lives on Instagram, TikTok, and Snapchat far more than on the website. A tool that analyses domains and ignores social reads as a tool built by someone who hasn't worked here.
The roadmap treated SWOT analysis as a practitioner feature. Participants treated it as something they produce for clients and rarely consult themselves. Same artifact, entirely different job — which changes where it belongs in the product and how much polish it needs.
User 1 was explicit about not trusting AI-generated figures he couldn't trace. This turned source attribution from a nice-to-have into a structural requirement: every number in the output has to show where it came from, or the report is worse than useless — it's a liability in front of a client.
Tool stacks were completely fragmented — four practitioners, four different setups, no overlap worth calling a standard. The one constant: ChatGPT is the fallback and the mental yardstick. Roya Plus isn't competing with SEMrush in users' heads. It's competing with "I could have just asked ChatGPT."
Five changes came out of synthesis. The two that mattered most:
The original hero output was the analysis report. But a report answers "how am I doing?", and the interviews kept surfacing a sharper question underneath it: how much do I need to spend to compete?
So the product's headline output became a calculated marketing budget — a concrete figure, derived from competitor spend estimation, with its working shown. "SAR 18,500/month to compete in your category." That's a number a business owner can act on and a practitioner can defend. It converts the same underlying data into a decision instead of a description.
MENA and Arabic-first weren't where this started; they're where the evidence landed. That distinction matters for how the product gets positioned — as a market gap the research identified and validated, not a demographic the team picked in advance and then justified. It's also a more honest story, and honest stories survive investor questions better.
There's a genuine tension in the findings, and flattening it would have been the easy call. Practitioners are the users. Business owners are the commercial target — the ones with budget, and the reason lead generation is the business model at all.
Designing purely for practitioners produces a tool the buyer can't evaluate. Designing purely for owners produces a tool the operator won't adopt. I wrote this up as the central strategic debate rather than picking a side and presenting it as settled, because the resolution is a go-to-market decision the founders need to own — not a research finding I could hand them.
The recommendation: practitioner-grade interface, owner-legible output. The person driving it gets depth; the person receiving it gets a number and a reason.
Documented explicitly rather than buried, because pretending research answered more than it did is how teams get confidently lost:
Report language by country — Arabic or English isn't uniform across the region and four interviews can't map it. Pricing and willingness to pay — untested. Senior practitioner behaviour — the sample skewed hands-on; decision-makers may work differently. Snapchat and WhatsApp spend estimation — methodologically unsolved, and both matter enormously in this market. Offline-first SMEs — a real segment the current model doesn't serve.
The most valuable output of this project was a set of things the team stopped planning to build. Four conversations, run before the roadmap hardened, moved the user, the hero output, the market framing, and the trust model.
None of that shows up as a screen in a portfolio. It shows up as a product that had a reason to exist before anyone drew it.