Outbound systems
Cold Outreach Systems for MENA B2B — Email, WhatsApp, and a Reply Cockpit
We do not sell lists or blast templates. We build the machine that finds the right companies, makes something genuinely useful for each one, delivers it on the channel they actually read, and puts every reply in front of a human who decides what happens next.
Why generic cold email fails in this region
The playbook exported from US B2B SaaS underperforms badly in Egypt and the Gulf, for reasons that are structural rather than cosmetic.
- Email is not where decisions happen. Owners and ops managers live in WhatsApp. A cold email to a company info@ address is frequently read by nobody with authority.
- Trust is relational. "Book a 15-minute call" from an unknown foreign-looking sender is a weak ask. Showing up with something already built for them is a strong one.
- Language. Arabic-first businesses read English pitches as spam by default. A message in the recipient's language, phrased for their sector, changes the reply rate materially.
- Volume is not the lever. Regional markets are small enough that burning the addressable list with a bad campaign is a permanent, not a temporary, cost.
So the system we build optimises for relevance per prospect and for channel fit — not for messages sent.
The system, component by component
- Sourcing. A defined ICP turned into a real list from maps data, directories, trade registries, portals and public listings — not a bought CSV of unknown provenance.
- Enrichment and verification. Decision-maker identification, email verification, WhatsApp-reachability checks, and hard deduplication against everyone you have ever contacted.
- Per-prospect asset generation. The step that makes the whole thing work: something specific is produced for each prospect before any contact. For our own outbound that is a complete bespoke demo website; for a client it might be a tailored audit, a costed proposal, or a data snapshot of their market.
- Sending. Email through your own authenticated domains and WhatsApp through a gateway, throttled, batched, and rate-limited per day and per domain.
- Reply ingest. Both channels funnel into one place, threaded per prospect, with sentiment and intent classified on arrival.
- AI-drafted replies. A draft is prepared using the full context of the thread and the prospect's business.
- Human approval gate. Nothing is sent to a prospect without a person approving it — including follow-ups. This is the non-negotiable part of the design.
- Tracker dashboard. Every prospect's state — generated, sent, delivered, opened, replied, meeting booked — visible in one view, with the whole history per lead.
Deliverability engineering
Most "outreach automation" projects fail on infrastructure, not copy. What we set up as standard:
- Separate sending domains, so a mistake never touches your primary corporate domain.
- SPF, DKIM and DMARC configured and verified — with DMARC actually monitored, not just present.
- Warm-up before volume, then a per-mailbox daily ceiling that stays deliberately conservative.
- Throttling and randomised send windows in local time zones. Nothing sends at 3am Cairo time.
- Hard batch-size guards in code. We added ours after an over-send incident of our own, described in the outreach machine case study — the guard exists because we needed it.
- Bounce and complaint handling that suppresses addresses automatically and permanently.
- WhatsApp pacing that respects the platform's limits, because a banned number costs more than a slow campaign.
What we do not do
- No spray-and-pray. If the asset for a prospect is not genuinely specific to them, the prospect does not get contacted.
- No scraped garbage lists, and no data of unclear provenance.
- No fully automatic sending. A human approves outbound messages and replies, always.
- No fake personalisation tokens — the "I loved your post about {topic}" pattern is transparent and it costs you the account.
- No guaranteed-meetings promises. Anyone quoting you a reply rate before seeing your offer is guessing.
Proof: we run this on ourselves
The outreach system described here is the one that produced the conversation you may be having with us right now. It generates a complete, bespoke demo website for each prospect, passes it through an automated visual QA gate, delivers it over WhatsApp and email, and routes every reply into a co-pilot where a human approves the response.
Read the full case study, including the over-send incident and the guard we built afterwards.
Engagement model
Build — from USD 2,500 one-time. Sourcing, enrichment, asset generation, sending infrastructure, reply cockpit and tracker, handed over and documented.
Managed — from USD 1,000/month. We run the machine, you approve the replies and take the meetings.
Deliverability-only setup — from USD 900. Domains, authentication, warm-up and monitoring, if you already have a sending motion.
Third-party costs (data providers, mailbox hosting, WhatsApp gateway) are passed through at cost, on your own accounts, so you see them directly. Ad-hoc work outside the managed plan is USD 45/hour.
Frequently asked questions
Is cold outreach legal in Egypt, Saudi Arabia and the UAE?
B2B outreach is generally permissible across the region, but the rules differ by country and are tightening — Saudi Arabia’s PDPL and the UAE’s data-protection law both create obligations around personal data and consent, and unsolicited commercial messaging is separately regulated. We build honest sender identification, a working opt-out and suppression list, and per-country pacing into every system, and we recommend you take local legal advice on your specific campaign.
Do you provide the lead list?
We build it with you from a defined ideal customer profile, using public and licensed sources, then verify it. We do not resell purchased lists, and we will not run a campaign on a list whose origin nobody can explain.
Can you run outreach in Arabic?
Yes. Arabic-first sequences, written natively rather than machine-translated, with replies handled in the language the prospect used.
What reply rate should we expect?
We do not quote one before seeing your offer and market, and you should distrust anyone who does. What we will commit to is a system where every message is specific enough that a reply is a reasonable outcome, and where the numbers are measured honestly per campaign.
Will this get our domain blacklisted?
That is exactly what the sending-domain separation, authentication, warm-up, volume caps and suppression handling are for. Your primary corporate domain is never used for cold sending.
Can we approve every message before it goes out?
You do not just get to — you have to. Human approval before send is built into the design and is not an optional setting.
Tell us who you sell to
Send us your ideal customer profile and your current outbound results, if you have any. We will map what a system for your market would look like and tell you whether it is worth building.