Outbound for a software company with no sales team

A software company with four engineers and nobody whose job is selling decides to try outbound. Someone exports a list of six hundred companies, writes one message about custom software and digital transformation, and sends it to all of them inside a week. Two people reply, both asking to be removed. The company concludes that outbound does not work in this market. The conclusion is wrong. What failed was a list nobody had looked at and a message that could have been addressed to any company on it.

Outbound for a technical company is a research activity that ends in a short message. At the volume research allows, fifteen to twenty approaches in a week, it produces conversations. At the volume a mail merge allows, it produces silence, plus a sending domain that starts landing in spam folders. The work is almost entirely in the part that happens before anything is sent.

The list is built from an observation

A purchased database gives you a company name, an activity code and a phone number. None of that tells you whether the company has the problem you solve, so a message written from it can only be generic. The alternative is slower and it works: you build the list one company at a time, and a company enters it only because you saw something specific about how it operates.

Most businesses run their operations in public without meaning to. What you are looking for is an operational detail visible from outside:

• a retailer taking orders in the comments under its own posts, retyping the price and the delivery cost in every reply
• a clinic whose page asks patients to call between ten and twelve to book, because the appointment book is a paper diary on the reception desk
• a distributor whose invoice, when a delivery arrives at your own office, is a handwritten carbon copy
• a wholesaler who publishes the weekly price list as a photograph of a printed sheet
• a company advertising three data entry vacancies, which usually means figures are being retyped from one system into another
• a chain with four branches, four phone numbers, and stock counted separately in each

The sources are ordinary: the company's own pages, reviews on Google Maps where customers describe the failure in their own words, job advertisements, supplier and tender notices, and for local trade an afternoon in the district itself. In Egypt that means walking through a wholesale market or an industrial zone with your eyes open rather than reading about it.

One condition before a name goes on the list. The observation must correspond to something you have already built and operate today. If it does not, you are pitching an idea, and the conversation will end at the first technical question. This is where a defined scope pays for itself, and it is the practical argument in choosing what you are for.

Budget fifteen to twenty minutes per company. You should finish with three lines written down: the detail, what it plausibly costs them, and the name of the person for whom it is a problem. Fifteen to twenty researched approaches a week is a complete outbound programme for a company with no sales team.

The detail you noticed is the message

The first sentence has one job, which is to prove you looked at this company before you wrote to it. "We provide custom software solutions for retail businesses" proves nothing and could have been sent to nine hundred recipients, which is exactly how it reads. "You are taking around thirty orders a day in the comments and writing the delivery price out again in each one" proves that a human being spent time on them.

Fake personalisation is worse than none. A merge field with the company name in the greeting, or a line about how much you admire their work, is recognised instantly by anyone who receives outreach, and it costs you the credibility of every sentence that follows. The detail has to be one you could not have known without looking.

The second thing the message needs is the cost of that detail, stated in units the reader already uses. Hours of a person's day, orders that arrive after closing and are answered the next morning, two staff who spend the first hour reconciling yesterday. Do not state a percentage you cannot support. An owner who runs the operation himself knows what it costs better than you do, and a fabricated figure tells him you are guessing.

A message that could be sent unchanged to a second company was not worth sending to the first.

Four sentences and one small ask

The structure that works is short enough to read on a phone without scrolling.

1. What you noticed, in one specific sentence with no adjectives in it.
2. What it costs them, in one sentence, in their units.
3. One sentence of evidence that you can do this: a system you built that already does it, described by function and scale. Turning delivered work into a line like that is its own discipline, covered in proof that sells.
4. One small ask, answerable yes or no in five seconds.

Under 120 words in total. No attachment. A company profile PDF in a first message signals that the sender had nothing specific to say and expects the reader to go and find it. No calendar link, and no request for an hour of anyone's time. Nobody has an hour for a stranger. Ask for a reply, or for permission to send one more thing.

Before

"Dear Sir/Madam, greetings. We are a leading software house specialising in digital transformation and custom solutions, with a team of highly qualified experts across web, mobile and enterprise systems. Kindly find attached our company profile for your kind review. We would be delighted to schedule a meeting at your convenience to discuss how we can add value to your esteemed organisation."

After

"Mr Hesham, I noticed that orders on your page come in through the comments, and that the delivery price gets written out again in each reply. On a day with thirty orders that is most of one person's morning, and anything that arrives after midnight waits until the next day. We built and still operate the ordering system for a home appliance retailer in Cairo that takes orders on the same channel and writes them straight into stock and delivery. I can show you the two screens that matter. Would it help if I sent them?"

The difference is not politeness or tone. The second message could only have been sent to one company, and the ask is small enough that saying yes costs nothing.

Channel and language follow the buyer

The channel is not a preference. It is a fact about where the person who has the problem actually reads things.

Language follows the same rule. Write in the language the company's own public material is written in. A business whose posts are in Arabic gets Arabic, formal but plain, with the English technical term where the English term is the word people actually use. Mixing the two is normal in this market and reads as competence. What reads badly, in either direction, is a message that was written in one language and translated into the other.

Follow-up is a schedule, and it ends

Three touches over about three weeks. Each one carries something the previous one did not.

The first message is the observation, the cost, the evidence and the small ask. The second, five to seven working days later, is not a reminder: it is the thing you offered, or a second observation about the same operation. The third, roughly ten days after that, closes the loop and makes refusal cheap. Something close to: "I will leave this here. If the ordering side is not a priority this quarter, tell me and I will stop."

"Just following up" and "bumping this to the top of your inbox" add nothing and teach the reader that your messages can safely be ignored. After the third message, stop. Move the company to a slow list you revisit on a trigger: a new branch, an operations or finance hire, a regulatory change such as a new e-invoicing requirement, the run-up to a season. A trigger earns you a new first message. It does not earn you a fourth follow-up.

Reply rate and meeting rate, not open rate

Open rate measures whether an image loaded. It does not exist on WhatsApp or LinkedIn, and on email it has become unreliable now that mail clients fetch images on the recipient's behalf. Track four numbers a week instead:

• messages sent, counting only individually researched ones
• replies of any kind, including refusals
• conversations booked
• proposals that came out of those conversations

Refusals count as replies because a refusal proves the message was read and understood, which is more than a delivered mail merge ever proves. From experience rather than any study: fifteen to twenty researched approaches in a week typically produce a small handful of replies and one conversation, and some weeks produce none. The same effort spread across five hundred addresses produces nothing at all. When a reply does arrive, outbound is finished and something else begins, because turning an inquiry into a scoped project is a separate skill covered in the discovery call.

What compounds

Researched outbound leaves a residue that bulk sending does not. A company that says no still knows what you build and who you are, and people move, budgets open, and a name that arrived attached to a precise observation gets remembered when the problem finally becomes urgent. Six months of fifteen to twenty careful approaches a week is a few hundred companies who have met you properly. Six months of blasting is a burned domain and a reputation you will spend longer repairing than you spent building the list.

If a specific system is closer than another round of messages, talk to us about scoping it. If it is not yet, the more useful next step is deciding what you are for, before deciding who to write to.