Today's brief
Daily competitor & review intelligence — 2026-09-24
Today's pool was mostly new: Royex Technologies (Dubai), Leobit, Relevant Software and Brainhub (global mid-tier), and Algoriza (Cairo; Apps Square was checked first but had no third-party reviews we could read). Robusta Studio came back as the repeat for trend. Sourcing was still weak. Clutch returned HTTP 403 on 6 of 7 profile fetches (only Robusta's empty profile loaded), and G2, Trustpilot, DesignRush and The Manifest were also blocked. So most ratings come from GoodFirms and Techreviewer pages we fetched directly, with Clutch figures taken from search snippets only. Three findings matter today. First, Leobit prices its Build-Operate-Transfer (BOT) offer per engineer against the client's home market: $5,500 a month for an intermediate developer versus $9,500 in-house in the USA, and $7,200 for a senior versus $14,000. That is a pricing format Devya can copy for GCC buyers. Second, the most useful complaints came from the firms with the best scores. Royex's clients wanted better handover documentation, a post-launch roadmap and analytics. A Vodafone reviewer of Robusta asked for 'more proactive bug reporting' and a tech lead who is 'more present in our discussion with product'. Third, the only client-confirmed AI speed number in the pool is Relevant Software's '35% faster feature delivery', found in a Clutch review summarised by Techreviewer. Robusta, by contrast, has rebranded around 'AI-driven ecosystems' and still has no Clutch or GoodFirms client reviews at all, the same as every check since August. Counter-signal: Royex's GoodFirms score of 4.9 from 38 reviews includes at least seven 5.0 reviews posted between 27 April and 7 May 2026, all with near-identical wording. That reads as a review campaign, not independent feedback.
Do
Offer Build-Operate-Transfer to GCC clients who eventually want their own team, and publish a per-engineer monthly cost table (intermediate and senior) set against what the same engineer costs in-house in Riyadh or Dubai. Put the transfer terms (timing, fee, which staff move over) on the same page.
A GCC buyer weighing outsourcing against building in-house needs a side-by-side number and a clear exit. BOT sells the capacity without making procurement worry about lock-in.
Leobit's BOT page says: 'Monthly cost of hiring an Intermediate Developer is $5,500 through Leobit BOT (including salary and administration fee), compared to $8,500 in the UK, Switzerland, and Nordics, and $9,500 in the USA', and $7,200 for a senior versus $12,000 and $14,000. None of the regional firms reviewed today publish a comparable offer.
Measure what AI-assisted development actually saves on each engagement: record cycle time per feature before and after the AI workflow, report it in sprint reviews, and ask the client to put that number in their review.
Every agency now claims AI. A speed-up the client states in a third-party review is the only version a sceptical buyer believes, and it lets Devya charge for the gain instead of just absorbing it.
Techreviewer's summary of Relevant Software cites a review documenting 'AI-assisted development workflows enabling 35% faster feature delivery'. Leobit's reviews cite document creation 'reduced from hours to minutes via AI automation' and a 70% processing-time improvement. Robusta has repositioned as 'AI-driven ecosystems', but none of its five Sortlist reviews mentions AI.
Make the handover pack a priced line item that the client signs off before the final invoice: an architecture decision record, runbooks, an inventory of environments and secrets, a deployment guide and a list of known issues.
Documentation is the one soft spot that shows up even in near-perfect review profiles. A client who can't run the system without Devya either feels trapped or leaves angry.
A Royex client on Clutch: 'Documentation could have been more detailed, especially for the AI model training process and internal handover'. Techreviewer says Relevant Software's only 'minor areas for improvement relate to documentation quality rather than core delivery'.
Include product analytics setup in every launch, plus a 30-day readout for the client covering activation funnel, retention cohort, crash-free sessions and the three screens where users drop off. Then offer the ongoing version as a paid monthly tier.
Clients judge a launch by what happened after go-live. The numbers Devya hands them are the numbers that end up in their review, and the readout is a natural lead into a retainer.
A Royex client on Clutch: 'Providing additional post-launch analytics support would be one of the areas to improve ... more insights about customer behavior inside the app would have been additional value'. Brainhub's reviews quote post-launch figures such as '99.9% bug-free', '180 users and no crashes' and '1.17M app impressions / 3,200 active users'.
Put the tech lead on the standing agenda for every client product discussion, not only the PM or account manager, and say so in the SOW as a named attendance commitment.
Product decisions made without engineering in the room come back as rework and missed dates. Clients notice when the person who understands the system is missing.
Robusta Studio's Vodafone reviewer on Sortlist (3.75/5) praised the account manager but asked: 'I think we need the tech lead to be more present in our discussion with product'.
Report bugs to the client before they find them: send a weekly defect log (found, severity, status, owner) from the first sprint, well before UAT, and report bugs Devya caught itself separately from bugs the client caught.
A bug the client finds first costs trust. The same bug reported by Devya first earns it. The found-by-Devya share is also a quality number worth showing in sales.
Robusta's Vodafone reviewer: 'The main area for improvement lies in QA/QC: there's a need for more proactive bug reporting and ensuring tech team is in alignment with design and product'. Brainhub's reliability score rests on reviews calling its deliverables 'near bug-free'.
Collect reviews on every platform bots and aggregators can read, not just Clutch: GoodFirms, the Google Business Profile for the Cairo office, and Sortlist. Rotate which one each client is asked to use.
Clutch is blocking automated readers, and aggregators and AI assistants fill the gap from other sources. A Clutch-only review base is becoming invisible to the tools buyers now use for shortlisting.
Clutch returned HTTP 403 on 6 of 7 profile fetches today. Techreviewer builds its score by pooling Google Maps reviews: 27 for Brainhub (4.7) and 18 for Relevant Software (4.8). Brainhub has 53 Clutch reviews but only 1 on GoodFirms. Robusta's only readable client reviews are the 5 on Sortlist.
Add team-stability numbers to the site's trust bar and to proposals: engineer retention rate, share of senior engineers, and average time an engineer stays on one client account.
What clients fear most from an agency is the team changing mid-project. A retention figure answers that before it's asked, and it is backed by an employer reputation that buyers can check.
Relevant Software's homepage leads with '96% employee retention' and '92% senior talent' next to '98% client satisfaction'. Brainhub's Glassdoor is 4.7/5 with 100% recommending, compared with Robusta Cairo at 3.6/5, where employees cite 'low salaries, management issues' and overtime, and a Royex Glassdoor review titled 'Worst Company Ever'.
Avoid
Don't run review drives. Never ask a batch of clients for reviews in the same week with a template, and never accept a review that has no project detail. Spread requests across delivery milestones and ask for one specific number or trade-off in each.
A burst of near-identical 5.0 reviews is a pattern buyers and GoodFirms' own AI summary can see. It raises the score but gives a prospect no reason to choose you, and it invites doubt about the whole profile.
Royex's GoodFirms profile (4.9, 38 reviews) shows at least seven reviews posted 27 April–7 May 2026, each scoring 5.0 on every sub-score, with interchangeable wording ('well organized', 'steady', 'clear communication') and no project metrics. Only 1 of the 38 reviews is tagged 'Value for Money'.
Don't let directory profiles contradict each other or the website. Check founding year, headcount, office list and rate band on every platform each quarter, and keep one source of truth.
Procurement teams and AI summarisers compare profiles across sites. A contradiction reads as carelessness at best and padding at worst, and it is the first thing a sceptical buyer points out.
Robusta shows 2007 as its founding year on Clutch but 2012 on GoodFirms and Sortlist, lists 50–249 staff on directories against '200+' on its own site, and shows 2 hubs on Clutch/GoodFirms against 5 on Sortlist. Relevant Software lists $25–$49/hr on GoodFirms but $50–$99/hr on Clutch.
Don't reposition the brand around AI before at least one client review says, with a number, what AI did for them.
An AI rebrand with no evidence behind it makes Devya look like everyone else. A buyer who clicks through and finds nothing about AI in the reviews discounts the rest of the pitch too.
Robusta Studio is now marketed as an 'RTG Engine' of 'AI-driven ecosystems', while it has zero Clutch and GoodFirms client reviews and five Sortlist reviews that don't mention AI. By contrast, Relevant Software's AI claim is a client-reported 35% delivery speed-up.
Don't quote a rating in a proposal, deck or listicle unless the buyer can click through and verify it on the platform itself.
An unverifiable rating is worse than none: when the buyer checks and finds nothing, every other claim in the proposal loses credibility.
A 2026 'Top 10 Software Development Companies in Saudi Arabia' listicle cites 'Clutch Rating of 4.3/5 and Google Reviews of 4.2/5' for Apps Square, but its Clutch profile shows no readable reviews and the listicle's author ranks a rival firm above it. Algoriza's only client praise is on its own website.
Don't make urgent client issues wait for the next shared working hour. Publish an urgent-issue path with named on-call hours and a response time for clients outside the Gulf time zone.
Time-zone friction is the one complaint that survives in otherwise perfect Eastern European profiles, and it shows up exactly when the client is under pressure.
Leobit's reviews repeatedly mention time-zone friction slowing responses to urgent requests. Techreviewer notes that a Relevant Software reviewer 'flagged time zone overlap limitations'.
Don't hand approved designs to engineering without a joint feasibility review with design, product and the tech lead before anything is built.
When design, product and engineering drift apart, clients see it as bugs, rework and shifting dates, and they blame the delivery team, not the design.
Robusta's Vodafone reviewer praised the design team as 'highly responsive' but called for 'ensuring tech team is in alignment with design and product' as the main improvement.
Don't ship an AI feature as a black box. Every model or LLM feature goes out with its evaluation set, its prompts or training configuration, and a written procedure for retraining or changing prompts that the client's team has tried themselves.
Once Devya leaves, an AI feature the client can't evaluate or retrain slowly gets worse. The client blames the vendor for the decay, and that complaint lands in a public review.
A Royex client whose project included AI analytics wrote that documentation 'could have been more detailed, especially for the AI model training process and internal handover', and that 'a clearer roadmap for post-launch support would've helped plan better for the long term'.
Competitor sentiment
Royex Technologies
Web/app/e-commerce + AI agency · Dubai HQ, Qatar sales office, India + Bangladesh delivery hubs · founded 2013
- Structured communication: 'maintained clear communication through project management tools and weekly check-ins, with milestones generally being met on time' (Clutch)
- Proactive about changes: 'suggested adjustments were met with proactive solutions rather than delays' (Clutch)
- Broad delivery across verticals: fintech app with payment scheduling and a merchant dashboard, e-commerce with ERP and payment-gateway integration, an AI analytics dashboard for a food-delivery platform (Clutch)
- GoodFirms' AI summary cites 'Problem Understanding' in 32 of 38 reviews and 'Communication' in 28
- Handover documentation too thin: 'Documentation could have been more detailed, especially for the AI model training process and internal handover' (Clutch)
- No post-launch plan: 'a clearer roadmap for post-launch support would've helped plan better for the long term' (Clutch)
- Clients wanted post-launch analytics: 'more insights about customer behavior inside the app would have been additional value' (Clutch)
- Review pattern looks solicited: at least seven 5.0 GoodFirms reviews between 27 April and 7 May 2026 with templated wording and no project metrics; 'Value for Money' is tagged in only 1 of 38 reviews
Takeaway: Royex's own clients describe where the project ends: handover, analytics and the post-launch plan. Devya should sell exactly that as a scoped service, and win against the Dubai-front, South-Asia-delivery model on same-time-zone, Arabic-first delivery rather than on price.
Leobit
Custom software / SaaS / .NET + AI · Lviv delivery with Austin, Tallinn, Kraków, London offices · 170+ engineers
- Measurable AI results: '70% processing time improvement for a trichoscopy app' and document creation 'reduced from hours to minutes via AI automation' (Techreviewer summary of reviews)
- Feels like part of the client's team: clients across time zones 'consistently describe Leobit as feeling like an embedded team extension', citing daily standups and Slack (Techreviewer)
- On-time delivery 'across dozens of engagements' using Scrum and Jira (Techreviewer)
- Value for money: clients report significant savings versus onshore resources while keeping quality (Clutch, search snippet)
- Time-zone friction slows turnaround on urgent requests, a minor theme that recurs across reviews (Clutch, search snippet)
- A 2018 engagement 'noted a rocky start before staffing was resolved' (Techreviewer)
- Only isolated weak spots: one 4.0 schedule sub-rating and a few sub-4.0 cost/quality sub-ratings; Techreviewer finds 'no systemic weaknesses'
Takeaway: The format to copy is BOT priced per engineer against the client's home-market cost ($5,500 vs $9,500 for an intermediate developer in the USA). Rebuilt with Riyadh and Dubai benchmarks, it gives GCC buyers a reason to build their team in Cairo.
Relevant Software
Custom software + dedicated teams · Ukraine/Poland delivery, US-facing · 50–249 staff · fintech, SaaS, healthcare
- Communication is the most-praised trait: 'They were open, proactive, and responsive at every stage... Their team truly felt like an extension of ours.' (GoodFirms, fintech mobile app review, Oct 2025)
- Long relationships: 'I'm working with Relevant's team for more than 5 years already, we have delivered more than 9 software solutions.' (GoodFirms)
- Speed and price: 'very good service at a very competitive price... their speed of execution and responsiveness to change requests' (GoodFirms)
- Client-confirmed AI result: a review documents 'AI-assisted development workflows enabling 35% faster feature delivery' (Techreviewer summary of a Clutch review)
- Techreviewer: 'minor areas for improvement relate to documentation quality rather than core delivery'
- One reviewer 'flagged time zone overlap limitations' (Techreviewer)
- Rate band differs by directory: $25–$49/hr on GoodFirms vs $50–$99/hr on Clutch
Takeaway: The only firm in today's pool whose AI claim comes from a client, with a number. Devya should build the same evidence by measuring AI speed-ups per engagement and asking clients to state them in reviews.
Brainhub
Software house · Gliwice, Poland · 50–249 staff · JS/TS, Node, React/React Native, .NET · site now says 'part of STX Next'
- Partner, not vendor: clients praised Brainhub's 'be our partner' attitude compared with the usual client–vendor relationship (GoodFirms / search snippet)
- Quality results in reviews: 'a 99.9% bug-free mobile app and an app with 180 users and no crashes' (Techreviewer, from Clutch reviews)
- Quantified outcomes: 5 million monthly users and a 25% conversion increase (Techreviewer)
- Delivery discipline: on time and on budget, with milestones tracked in Azure DevOps, Jira and Confluence (Techreviewer)
- Techreviewer: 'No systemic or recurring weaknesses were identified'. The only soft spot is one older review with a 3.5 schedule sub-rating
- A client framed price as a future risk: 'Their hourly rates are typical for Polish standards, but in my opinion, it will go higher because of the good quality which they deliver' (Clutch, search snippet)
- Reviews sit almost entirely on Clutch (53) with only 1 on GoodFirms, so the profile is thin anywhere Clutch can't be read
Takeaway: Brainhub shows what a review base looks like when it is built on post-launch numbers (crash-free, bug-free, user counts). Devya gets the same effect by handing every client a 30-day analytics readout they can quote.
Robusta Studio
Digital agency / enterprise transformation · Cairo HQ; Sortlist and group site list Egypt, Saudi Arabia, UAE, UK, Germany · now marketed as an 'engine' of Robusta Technology Group (RTG)
- Account management: 'My account manager consistently manages the workload efficiently and communicates seamlessly with all teams' (Sortlist, Vodafone digital projects reviewer)
- Design and product leadership: 'The Product Team has strong leadership in guiding the tech team, while the Design Team is highly responsive and delivers good quality designs' (Sortlist)
- Seen as a partner in retail/B2B: 'considered a partner and business driver in our operation in Retail and B2B sectors' (Sortlist, Raya IT director)
- Enterprise client list on GoodFirms (P&G, Mondelēz, Vodafone, Orange, Pepsico, AXA, Fawry)
- QA: 'there's a need for more proactive bug reporting and ensuring tech team is in alignment with design and product' (Sortlist, 3.75/5 review)
- Tech lead absent from product discussions: 'we need the tech lead to be more present in our discussion with product' (Sortlist)
- Still no client reviews on Clutch or GoodFirms, unchanged at every check since August, while it charges the highest rate band in the Cairo rotation
- Directory data contradicts itself: founded 2007 (Clutch) vs 2012 (GoodFirms/Sortlist); 50–249 staff vs '200+' on its own site; 2 hubs vs 5
Takeaway: Trend holds: Cairo's premium agency is betting on an AI/enterprise rebrand and GCC offices rather than on reviews. Its one detailed enterprise review names QA reporting and tech-lead presence as the gaps, which is exactly where Devya should pitch Egyptian enterprise accounts.
Algoriza
Custom software + staff augmentation · Cairo · 11+ years (Apps Square, Riyadh/Cairo, was checked first and dropped: no readable third-party client reviews)
- On its own site only (not independent): 'distinguished by its good selection of technical teams, following up with them, and maintaining continuous development for them'
- On its own site only (not independent): praise for 'phenomenal support and dedication' on digitising Egypt's nursing syndicate
- No independent client reviews on any platform we could read, so a buyer has nothing third-party to check
- Apps Square has the same gap: its Clutch profile has no readable reviews, and the only rating in circulation ('Clutch 4.3/5, Google 4.2/5') comes from a Saudi SEO listicle written by a rival, not from the platforms
Takeaway: Cairo's review vacuum is wider than one firm: Robusta, Algoriza and Apps Square all compete for the same MENA buyers without a readable independent review base. The first Egyptian firm with 15–20 specific, dated reviews on crawlable platforms leads by default.
Opportunities
- 01Cairo BOT for GCC: Leobit publishes a per-engineer BOT cost against US/UK in-house pay, and none of the regional firms reviewed today does the same for Riyadh or Dubai. A Devya BOT page with GCC in-house comparisons and clear transfer terms targets buyers who want an owned team eventually.
- 02Same-working-day delivery as a positioning line: Royex runs a Dubai/Qatar front office on India/Bangladesh delivery at $25–$49/hr, and Leobit and Relevant Software both get time-zone complaints. Cairo overlaps the Gulf working day almost fully, so Devya can sell overlap plus Arabic-first delivery rather than a lower rate.
- 03Post-launch analytics tier: Royex clients asked for it by name and Brainhub's review base is built on post-launch numbers. A 'launch + 90 days of insight' package (instrumentation, monthly behaviour readout, top-3 fix list) is a natural bridge from a project to a retainer that most shops leave unsold.
- 04Client-confirmed AI results: Relevant Software's 35% is the only client-stated AI speed-up in today's pool, while Robusta rebrands around AI with no reviews to back it. A few Devya reviews quoting measured AI gains would put Devya ahead of both.
- 05The Egyptian review vacuum persists: Robusta (no Clutch/GoodFirms client reviews at any check since August), Algoriza and Apps Square all lack independent reviews anyone can read. Reviews spread across GoodFirms, Google Business Profile and Sortlist would make Devya the most visible Cairo option to aggregators and AI assistants.
- 06Enterprise QA pitch in Egypt: Robusta charges $100–$149/hr, and its one detailed telco review scores it 3.75/5 and names proactive bug reporting and tech-lead presence as the gaps. A QA-led proposal (weekly defect log, Devya-found vs client-found share, tech lead in every product meeting) goes straight at that gap for Egyptian telco, FMCG and retail accounts.