How did Microsoft find customers?

The promised software had to work

Bill Gates’ retrospective says Paul Allen wrote to MITS claiming they had BASIC for the Altair, then followed up by telephone. MITS president Ed Roberts offered the deal to whoever first produced a working version. Allen demonstrated the software in Albuquerque after the founders built it without access to an actual Altair. [1]

Microsoft’s official history identifies MITS as its first customer in 1975 and records sales of $16,005 that year. This establishes a named early customer and historical revenue. The supplied material does not verify the precise first paid contract terms or an early prospect-to-customer conversion rate. [2]

A language made a new machine more usable

[Inference] The Altair supplied a highly specific initial context. A programming language had immediate relevance to what owners could do with the machine. The manufacturer also had a reason to help distribute that capability. This is a stronger partner rationale than simply seeking association with a prominent company: the combined offering addressed a recognizable customer job.

[Inference] The demonstration converted a technical promise into inspectable evidence, but the outreach should not be generalized into advice to sell unsupported capabilities. Its commercial meaning depends on what was subsequently delivered. The case documents an unusual founder-led sequence, not the reliability of speculative promises as a repeatable acquisition method.

One negotiation can reach many installations

The official 1981 timeline records IBM launching its PC with MS-DOS and other Microsoft language products. Microsoft identifies Steve Ballmer as its first business manager, hired in 1980, and credits him with helping land the IBM agreement. A company-hosted biography records Harvard mathematics and economics training and Stanford business-school attendance. [3][4][6]

[Inference] Hardware distribution can combine a difficult commercial negotiation with broad downstream access. It also creates dependence on the partner’s terms and customer relationship. The relevant economic analysis must include that concentration rather than treating every installed copy as an independently acquired user or assuming that distribution access automatically translates into active use.

The modern account needs adoption work

Current FastTrack documentation describes planning, onboarding and partner involvement for eligible subscriptions. Microsoft 365 guidance generally specifies at least 150 eligible licenses, with distinct conditions for migration benefits. These are modern service conditions and should not be projected backward onto the MITS or IBM agreements. [5]

[Inference] Organizational software adoption introduces migration, security, interoperability and user behavior into the purchase. Deployment help can turn licensed capability into work that people actually perform. Partners may extend delivery capacity, but ownership still matters: the customer needs clarity about responsibilities and the conditions under which a new workload is ready to expand.

[Inference] This support model also makes eligibility part of the customer experience. Benefits tied to particular subscriptions should be assessed as specific commercial commitments, rather than assumed to accompany every purchase.

Different businesses require different denominators

[Inference] OEM licenses, enterprise subscriptions and cloud consumption have different cost structures. Historical software distribution does not establish the cost of infrastructure or specialist adoption services today. The record therefore cannot support a synthetic company-wide CAC, lifetime value or retention estimate. The original sales figure is a dated fact, not a benchmark for modern economics.

[Inference] Microsoft’s transferable lesson is complementary value: make a partner’s offer more useful and develop the capabilities that help customers adopt it. A better expansion assessment would distinguish licensed seats, activated workloads and sustained use. That preserves the insight of the early distribution relationships while recognizing that a present-day enterprise contract is only one stage in a much more complicated customer journey.