Positioning + GTM systemData and AI engineeringNovi Sad, Serbia

SmartCat

From engineering depth to a go-to-market that sells itPositioning, three GTM axes, five ICPs, a demand engine and revenue-linked measurement.

SmartCat has built production-grade data and AI systems since 2015 and grew to around 85 people almost entirely on referrals. Between March and April 2026 we worked as its external go-to-market partner. We rebuilt the positioning, mapped who buys and why, packaged the offer, designed how demand gets created and measured, and handed the whole system to SmartCat's own team to run.

Industry: Data and AI engineering
Market: EU and UK enterprises and scale-ups
Engagement: External GTM partner, handed over to their team

What clients say

“The work gave our leadership and our marketing and sales teams one shared positioning and a documented go-to-market playbook, handed over to our team to run on its own.”

Nenad Božić

Nenad BožićCo-founder, SmartCat

Where SmartCat stood

Almost all revenue came from existing clients, and new logos came through the CEO's personal network. The referral network was the only channel that worked consistently, yet nobody had ever written it down. At kick-off we mapped it for the first time and found the nodes it actually ran through: a handful of people in San Francisco and the Netherlands, and two technology partners.

The founders already had a clear conviction: the Central European time-and-materials story was over, and the future was selling measurable outcomes. What they did not have was a commercial system that could carry that conviction to buyers who had never heard of them. The positioning sounded like every other engineering shop, the case studies were anonymous and heavy on technology, and the marketing numbers they tracked did not connect to revenue.

What made it hard

Our strategic diagnosis named the blockers plainly. There was no CRM, so the pipeline was invisible. There was no demand generation function. The person leading growth was carrying three or more roles at once. Thought leadership existed on paper only, the referral system ran on memory, and the positioning was new and untested. We also mapped the team's capacity against what the system would need, because a plan nobody has time to run is not a plan.

The irony was that strong proof already existed and was not being used. One client had agreed to share both risk and upside with SmartCat. Another had gone through the same model twice: paid discovery, a fixed-scope MVP, a retainer, and then a second project. Those were exactly the stories that sell outcomes, and they were sitting unused.

Positioning: eleven versions to one sentence

We started on 5 March 2026 with a working claim, "We move AI from pilot to production", and a four-week plan: positioning and ICPs in week one, a messaging matrix in week two, the offer and website in week three, the go-to-market playbook in week four. Every document had one review round and 48 hours for feedback, so the work kept moving.

The first claim did not survive. A competitive analysis in three tiers showed SmartCat sitting in the crowded middle, and "AI" alone put it next to everyone else. The positioning went through eleven versions until the sync on 31 March locked the line that stayed: "SmartCat is the engineering-led partner for data foundations and AI-enabled operations, built to last, not to demo." The tagline came from the same place: built on a decade of data engineering before the hype. We also wrote down what SmartCat would no longer call itself, including "end-to-end AI solutions", "AI-native" and "full-stack". The offer was organized into three axes: Data Foundations, Applied AI Systems and Governed Enterprise Workflows, each with its own statement per ICP.

What customers told us, and what it changed

In April we interviewed existing clients, and their answers corrected the messaging. Buyers did not describe their need as "AI transformation". They described ideas and problems that had been waiting for the right technology. One client said the thing that set SmartCat apart was that they listened first and built second. Another valued seeing where the work stood at every phase, and named that as the reason the cooperation lasted.

The interviews also showed where the proof was strongest: data engineering. So we recommended that every AI claim be reviewed before external use and tied to a concrete system SmartCat had already delivered. The story got smaller and more believable, and that was the point.

The offer and the demand engine

The offer started as a single linear path and became four entry points, so a buyer could come in where they already were: a workshop, a direct MVP, a consultation, or taking an existing MVP to production. Paid discovery stayed at the center, run by senior engineers, with one rule: no paid discovery, no delivery.

The internal pipeline sketch we received was built on outbound alone, and we pushed back on it. Outbound only captures demand that already exists. These buyers purchase roughly once a year, stay with a vendor for three to five years, and arrive with a shortlist of two or three names. If SmartCat is not on that list before the buying moment, outbound comes too late. So we added a mindshare layer: social selling from the founders and the sales team as the foundation, organic first, and ads only to amplify what already works. We also argued that engagement scoring was too heavy for a pilot of 50 accounts.

The demand engine that came out of it sells through engineering. It runs on a technical webinar built around a concrete question, ABM on LinkedIn with tiered account lists and the founders' posts amplified to them, messaging tested on LinkedIn before any outbound, and senior engineers in discovery calls. Sales steps in only when an account raises its hand.

The order of things, and what we handed over

The diagnosis set the sequence: foundation in April and May, validation in May and June, scale in the third and fourth quarter. Prerequisites come before amplifiers. The one thing that had to happen first was getting the person leading growth out of three jobs. Instead of a list of channels, we wrote seven hypotheses, from thought leadership and outreach to events, webinars, partnerships and referrals, each with a signal to continue and a signal to stop, at a pace of two launches per quarter. The first two tests were planned: a technical webinar on 29 April, which also tested the ABM setup and the Data Foundations proof, and a conference in London on 11 May.

Everything was handed over to SmartCat's leadership, marketing and sales teams to run on their own:

  • 17 documents, from the strategic diagnosis and positioning to the ICPs, proof mapping, messaging matrix and offer packaging
  • A discovery call playbook, channel architecture, a thought leadership system, a content distribution playbook and a 90-day content plan
  • An ABM process and an assessment of the commercial team with a gap analysis
  • A six-stage ROI model that replaces MQLs: revenue, pipeline, active focus, future pipeline, ICP cluster and brand metrics
  • A new case-study format built on business outcomes and named stories

What changed

Here the result is the strategy itself. A company that grew on referrals now has a written, repeatable way to win new clients on purpose: one positioning that leadership, marketing and sales agree on, an offer with clear ways in, a demand plan where every channel has to earn its place, and a scoreboard that follows the path to revenue, stage by stage. The engagement ended with the handover, and SmartCat's own team now runs the system.

Who we are

The people who build your whole commercial system and run it themselves.

Strategy, marketing, sales, customers and the technology under them, under one roof and with the same people from day one.

Funky Enterprises

The people who build your whole commercial system and run it themselves.