The class of question mid-market and family-owned businesses across India and the GCC ask before hiring a large IT organisation — and a working diagnosis of the alternative.
The board has asked for a digital strategy. A large consulting firm has offered to prepare one. A vendor has proposed a platform. The CEO — who runs a successful, 300-person, mid-market business without a chief information officer, chief technology officer, or head of transformation — is holding three PowerPoints and a quiet suspicion that none of them will answer the question actually being asked. The question is: what should this business, at its actual scale and with its actual constraints, do first? Not what could a Fortune 500 do. What should we do? The strategies on the table assume an in-house engineering team the business does not have and does not intend to build. The vendor proposal assumes a platform investment the business is not ready to authorise. The consulting proposal assumes a two-year programme the business cannot politically sustain.
The digital-transformation industry was built for a specific type of buyer — the large enterprise with an existing internal IT function, an existing capital budget for technology, and an existing appetite for multi-year programmes. Mid-market and family-owned businesses are a different buyer. Their constraints are different. Their governance is different. Their appetite for reversibility is much higher. Their tolerance for programmes that do not visibly earn their keep is much lower. And yet almost every service model, methodology and reference framework in the industry has been designed around the large-enterprise buyer. The mid-market buyer is being sold a version of the large-enterprise approach, and being asked to build the internal IT function that approach requires — even when that internal function is not what the business actually needs.
The cost of doing nothing is real — competitors modernise, buyer expectations move, and the business drifts. The cost of doing the wrong thing is larger. Every major transformation study — from HBR's most-cited pieces (Tabrizi et al., 2019) through McKinsey's long-running work on transformation shortfalls — reports failure rates that would be commercially unsupportable for any business with the buffer a mid-market operator has. For a mid-market business, the wrong programme is the difference between an acceptable year and a bad one. The specific costs show up as capital committed to a platform that never gets used properly, a team hired for a function the business does not actually need, and a leadership team that becomes cynical about digital.
Software vendors sell platforms. Their commercial interest is that platforms get bought. The question a mid-market CEO needs answered is not *which platform is right for us* but *do we need a platform-level answer at all, and if so, in which sequence*. The answer to that question is not in a software vendor's interest to give. Similarly, a systems integrator sells implementation capacity. Their commercial interest is that implementations get authorised. The question of *should we implement anything at this scale* is not in a systems integrator's commercial interest to raise.
The leadership team. The successful mid-market business already has a set of senior operators who run it. The digital work should extend their capability, not replace them. Also do not attempt to install an internal IT function of the size a large-enterprise transformation would require — the operating economics of the business will not sustain that headcount, and the political weather of the business will not sustain the accountability shift.
The relationship between the business's operating model and its systems. Most mid-market businesses do not need to be transformed. They need to be reengineered — a smaller word, a more precise one. Reengineering means: name the two or three operating-model constraints that are actually holding the business back, sequence the work to release them one at a time, and hold the operating rhythm while it happens. This is what John Deere did when it treated software and data as a first-class business rather than an IT function [see the case analysis at /analyses/john-deere-legacy-manufacturer-becomes-data-company]. The scale differs; the operating-model logic does not.
The systems the business already has. In almost every mid-market case, the answer to *what should we do first?* is *make what we already own work as one connected system*. The CRM, the ERP, the accounting system, the operational tools the team actually uses — most of them are individually adequate. The gap is between them, not inside them. Integration is unglamorous work. It is also the work with the highest return per rupee, dirham or dollar spent in a mid-market business.
The two or three workflows that consume the most human time and produce the highest error rate. Not everything. Not what a vendor's automation platform is optimised for. The specific workflows in this specific business. Once identified, automate them progressively — one workflow, exposed to the team, running for a quarter, refined, before the next one begins. This is how a mid-market business earns the credibility of the digital work before authorising the next round.
AI helps where it earns its keep — in the specific workflows above. It does not help as a general programme. The mid-market business should approach AI the same way it approaches any operational investment: what specific workflow, what specific outcome, what specific measurement. If AI cannot answer those three questions, the AI investment is not ready. The AI-readiness discussion in Jettifi's working paper at [/insights/ai-readiness-is-a-culture-problem] applies here directly.
The right sequence for a mid-market business is: diagnose in weeks, not months. Fix in quarters, not years. Extend in stages. The first quarter should produce a specific, visible operational improvement — the reconciliation that used to take a week now takes a day, the pipeline the CEO could not see now updates daily, the approval that used to take three days now takes three hours. That visible improvement earns the credibility for the next stage. A typical eighteen-month arc for a mid-market business: months 1–2 diagnose; months 3–6 first improvement; months 7–12 second improvement; months 13–18 architecture consolidation. Each stage self-funding.
The alternative to building a large internal IT function is a smaller model: senior digital leadership, engaged fractionally where full-time is not warranted; specialists assembled per engagement where deep expertise is required; a small permanent internal capability that owns the operating rhythm once the systems are stable. This is the model Jettifi is designed to be. The client does not build a large internal IT organisation. The client operates the reengineered business, with the right senior partner alongside for the periods where the reengineering is happening.
A Jettifi engagement of this class begins with a two-to-four-week diagnostic. Not a strategy document. A diagnostic. What are the two or three operating-model constraints actually holding the business back? What would releasing each one look like operationally? What is the smallest reengineering sequence that could release them? Senior Jettifi core runs the diagnostic; specialists are assembled around the specific sequence the diagnostic produces. The engagement is designed to end — the client owns the operating rhythm, the systems run as a connected whole, and Jettifi returns when the next reengineering question arrives. That is Digital Independence.
Discuss the two or three constraints actually holding the business back
Talk to Jettifi →For the underlying working paper, read the Digital Reengineering pillar or the executive edition of The Guide.