Case studies of working with TechMine

Fifth year in business, fifteen people, a stable team, consistent growth. From the outside, a success. The owner described it differently: “We do more and more and get less and less out of it.”

The portfolio contained a dozen or so types of projects, from mobile apps to systems for industry. Each one different, each requiring separate discovery, pricing and onboarding the team into a new domain. Nothing repeated, so nothing was cheaper the second time. Sales relied on referrals from the main business partner, and the owner personally took over every enquiry and closed every client.

We broke the previous year’s revenue down into projects and compared each one with the actual time spent by the team. Four projects produced more than half of the margin. Seven: practically zero. Three brought a loss nobody had counted, because fixes after acceptance were no longer logged to the project.

The pattern could be described in one sentence: the company earned money where it had already done something similar, and lost it everywhere something had to be done “from scratch”. In sales terms the company was doing well, because it relied on warm referrals, personal authority and the owner’s experience. The problem was that every sales opportunity that came in was taken on and, as far as possible, closed.

Segmentation based on profitability. The four profitable projects shared a common denominator: logistics companies, 100 to 500 people, with a specific type of system that needs an integration layer and further typical automations built on top.

A value proposition in the client’s language. Instead of “we build custom software”, a description of what changes when data stops being retyped by hand between systems.

Disqualification criteria. We wrote down when the company does not take on a project: no source system, no decision-maker in the conversation, no budget above a defined level.

Pricing based on the result. No more “how many people × how many weeks”.

The company rejects about half of the enquiries, most often during the first conversation. It runs fewer processes, but deeper ones. The team does not invent new solutions in every other project, and the same type of integration done once again is priced higher, costs less and the client gets a faster implementation than from competitors. At the same time, communication was narrowed to a specific segment, which brought new clients from outside the direct circle of referrals.

In twelve months turnover grew by 63%, profit fourfold, and the average project value by 82%. The number of projects fell.

The most important change has no indicator: for the first time the company can say why it won or lost a specific project, and the pressure to “close at any cost” is gone.

Segmentation pulled along decisions unrelated to sales: dropping two technologies used only in unprofitable projects and changing the way fixes after acceptance are billed (3 developers had to change their stack).

Thirty-five people, a full calendar and one person through whom every conversation with a potential client passed.

The owner had already tried twice. He hired a salesperson; they parted ways after six months. Then a second one; after four. The conclusion was: “In our industry sales cannot be delegated, because you have to know the technology.” It was convenient, because it removed responsibility. It was also untrue. The owner had periods of many weeks when he was personally involved in delivery and did not deal with sales at all; some leads were not followed up for fear of “clogging up production”, so the funnel dried up and the company started running out of opportunities.

We asked three people how sales worked: the owner, the previous salesperson and the person doing estimates. Each described it differently.

There was nothing to hand over, because the process had never been formalized. The previous salespeople did not leave because clients wanted a technical counterpart; they were given a target without a described route to it. The owner ran conversations well, but intuitively. And what cannot be described cannot be handed over.

Six stages with hard transition criteria. Not “the client is interested”, but: we know the mechanism of the problem, we know the cost of keeping it, we know who decides and what will influence the decision.

Packaging what the owner did intuitively. We implemented our standard process and added specifics based on the practice of the owner, who had been selling himself and had a few interesting tactics. The whole thing was written down, parameterized, filled with content at every stage, and measurement of activities was built on top. We filled the communication based on the Value Triangle. Before, the conversation started with the technology stack and the hourly rate.

Two people from the team run sales from the first contact to the proposal. The owner joins roughly every third process, when deeper technical expertise is required or when the buyer persona is a CTO. The weekly review takes forty minutes and is based on what is visible in the system. The company did not hire a single salesperson in that time.

The same situation, two years ago and today. Opening the first conversation:

The time from first contact to closing shortened by roughly a third, mainly because conversations stopped waiting for a free slot in one person’s calendar. Conversion from prospect to client rose from 20 to 40%.

The company stopped having a single bottleneck. The turning point came when, during the owner’s holiday, the team closed a highly technical client without his involvement.

Twenty-two people, eighth year on the market, no particular specialization, but the client profile was medium-sized and large companies with large projects.

To price such a project, the company did discovery: several meetings with the client’s team, a review of systems, mapping processes, an architecture proposal. Six to ten days of work by an analyst and an architect, before anything was signed, for free. The owner treated it as the cost of acquiring a client: “You cannot price it without this, and the client will not pay for something they have not bought yet.” That was the owner’s belief about how to start working with clients.

The analysts’ time spent on proposals did not go to any project; it simply vanished from the available capacity. Added up, it came to about a hundred days of work a year, forty of them on opportunities that ended in nothing.

Three situations repeated among the losses: the client postponed the project, disappeared without a word, or took the finished analysis to a vendor who could price it lower because they had something to compare against. The most important thing, however, was different: the free analysis set the relationship. The company appeared as a bidder courting a contract, not as an expert diagnosing the situation.

Separating the workshop as a distinct, paid service. Two days of workshop plus a week of own work, ending with a document that stays with the client regardless of whether they order the implementation.

Describing the result, not the scope. The client does not buy a “pre-implementation analysis”, but an answer to the question of whether the project makes sense, what it really costs and what happens if the company does not go ahead.

A price deliberately low relative to the project. Low enough to make the decision easy, and high enough to require approval from the person with the budget. The second point mattered more than the money.

A conversation about a project leads to a workshop, not straight to a proposal. A client who does not want to pay for it would almost never have ordered the implementation; the company finds that out after two weeks, not after six days of team work. When the client pays, they send the right people to the meeting.

The same situation, before and after. The response to a request for an estimate:

In nine months there were twelve paid workshops, nine of which turned into an implementation project, a significantly higher success rate than with the old free analyses.

About forty days of team work a year stopped being given away for free. The company also stopped competing with vendors who priced the project based on its own analysis.

Twenty-eight people, good deliveries, clients coming back for further projects. The company did not have a problem with sales, but with the model of working with clients.

Revenue looked like a saw blade: a good quarter when two large implementations started, then a quarter when half of the team was looking for something to do. Financially good periods covered the weaker ones, when the team had to be kept. Work was planned almost spontaneously, because nothing could be predicted. Maintenance of systems supposedly existed, but in a strange form: the client called when something stopped working, someone fixed it, some of the hours ended up on an invoice, some did not. Nobody counted how many there were.

Counting those scattered hours was the most uncomfortable moment of the engagement. The team spent over a tenth of its time on uncontracted development and repair work, and the company invoiced only part of it so as “not to spoil the relationship with the client”.

That was a symptom, though. The real problem lay in the timing: maintenance was raised after the implementation was accepted, when the client had just paid a large invoice, the project had formally ended and a conversation about more money sounded like an attempt to upsell. So it almost never happened. And when it did, the company described maintenance as a list of activities; nobody said what it protects against, so the client did not understand why they should decide on further spending.

Maintenance and development as a separate service, not an add-on.

Moving the conversation before the contract is signed. Maintenance became part of the implementation proposal, one of the stages on the timeline. That was the single most important change.

Describing it in the language of the client’s risk. Instead of a list of activities: what happens when a system that sixty people work on stops working on a Wednesday morning.

Recovering the old portfolio. A scheduled review conversation with every client from the last four years, not a sales call, but one about what has changed and what the risks are.

The maintenance conversation takes place before the project starts and clients treat it as a natural part of the proposal. Nineteen of twenty-three active clients have a contract; the four exceptions are deliberate decisions. The review conversations bring a side effect nobody planned: the company learns about upcoming projects before a request for proposal is created. Last year three implementations started exactly that way and never reached the competitive market.

The share of recurring revenue rose from about 10% to about 47%. Total revenue also grew, because setting clear rules of cooperation with clients and communicating value did not drive clients away, it tied them in more closely.

A quarter without a new large implementation is no longer a quarter in which work has to be found for half of the team.

A marketing agency, 8 years on the market, good work, strong competences, an impressive client portfolio. And a permanent feeling that the work is worth more than clients pay for it.

The first conversation always looked the same. The client asked about prices, packages, hourly rates, the owner gave a range, the client said they would be in touch. The agency had a transparent price list with three packages, and that was its biggest problem, although nobody saw it that way. Asked why a client should choose them, the answer was: “Because we have a lot of experience and a strong client portfolio.” Every competitor said the same.

The agency sold activities and the client bought activities. In such a conversation the only differentiator can be price, because two identical lists of services differ only in the amount at the bottom.

A price list given at the start framed the whole conversation in one dimension: cheaper or more expensive. Nobody ever asked the client how much their current situation was costing them. We checked the last nine conversations: in seven, the price came up before anyone had established what problems the client was really facing.

Reversing the order according to the Value Triangle. First the mechanism of the problem, then the result described as a change in the client’s situation, and only at the end the solution and the price.

Questions establishing the cost of the current situation. They give the agency material for pricing, and the client a reference point for the amount that comes later.

Taking the price list off the website. Not hiding prices, but moving the moment at which they appear.

Portfolio review. Parting with three clients on whom the agency was losing money.

The first conversation takes longer and does not end with sending a proposal, but with establishing whether there is anything to talk about. The proposal is shorter than before and clients comment on it more often instead of asking for a discount.

The same situation, before and after. A question about price in the third minute of the conversation:

Four months after implementation, the average service price was 55% higher. The offer, the team and the scope stayed the same; what changed was the conversation that leads to it.

The departure of the three lowest-priced clients did not lower revenue, because two new ones came in at the new rates in the same period. It did free up about 20% of the team’s time.

Twenty people, nine years on the market, a strong portfolio and no active sales activity.

The model worked for nine years, but stopped being enough. The number of referrals was not growing, because the owners’ network of contacts had its ceiling. Something else hurt more, though: what came from referrals was whatever happened to come. A quarter with three large enquiries, then a quarter with one small one. Once full-service marketing, another time organizing an event, and yet another time social media alone.

An earlier attempt to send emails to a purchased database brought no results at all. The conclusion was: “Cold mailing does not work for us.”

That was not proof that prospecting does not work, only that sending without segmentation and without a good reason for contact does not work. The deeper problem: the company could not say who its good client was. The answer was “it varies, we are flexible”, and you cannot look for someone you cannot describe.

Segmentation backwards, from completed projects. Three years of work analysed for margin, how the project went and whether the client came back. Two segments emerged.

Problems instead of lists of benefits. Not “all companies in industry X”, but companies in which we had observed a typical problem they struggle with.

Multichannel sequences. Coordinated contact by email and LinkedIn with different angles of approach, not the same generic content in two channels.

Division of roles and measurement. A separate person for the first contact; the owners come in only at the needs discovery stage. After a quarter we could measure specific numbers in a repeatable way.

Prospecting runs continuously, led by one person; it is part of the company’s work cycle, not a campaign that gets switched on. Referrals still provide most of the projects, but they have stopped being a constraint. Quarterly workload planning is based on a forecast derived from the volume and effectiveness of the prospecting being done.

Over a year, the number of projects won doubled. About 40% came from outbound, the rest still from recommendations, of which there were no fewer.

More important than the numbers was the change in dependence: the company stopped waiting for someone to tell someone about it.

Fourteen people, no salespeople, projects from several hundred thousand zlotys upwards. Clients put strong pressure on lowering the price, and the owner usually gave in.

The pain was not about the number of projects caught, but about the fact that despite a lot of work, the company had been recording a loss for many months. The team was fully booked, and yet the owner was putting money into the business from his own pocket. The company was recognized locally and had clients mainly from the area.

A review of projects showed that a large part of the team’s working time was not invoiced to clients. Many small clients were served at low hourly rates, generating a lot of activity on the company’s side. There was no active selling; mainly local inbound traffic was handled, because the company had operated on the local market for many years and had no significant competition. Despite that, clients put pressure on prices and stretched payments beyond the agreed terms.

Building a repeatable sales process, defining a clear client service process, from qualification to handover to the implementation team.

Deliberately letting go of some clients. Clients who did not accept the new rules of cooperation left (but some came back after a few months).

Rules for starting cooperation. Introducing subscriptions and prepayments for new clients.

Changing the task management system. Moving from Jira to ClickUp: close tracking of working time, logging to projects and a reduction in headcount.

There are 40% fewer clients and the team is smaller too, but measured client satisfaction with service has risen, the company is profitable and is starting to move beyond the local market.

On the local market the belief is that this company “is expensive, but good”.

The biggest change concerns something else, though: the owner stopped rescuing a sinking business at the expense of his own family’s financial security.

The company implemented ERP and WMS class systems. Preparing proposals was heavy and expensive: pre-implementation analysis, scoping, pricing modules, a schedule, often a workshop at the client’s site. Four to six days of a consultant’s work for one proposal.

Several dozen proposals were produced a year, and roughly one in six or seven was won. The internal rule was: if someone sends an enquiry, we prepare a response. “We cannot afford to let anything go,” the owner said. The numbers showed the opposite.

We counted the cost of preparing proposals. The company spent more on it than on any client acquisition activity, only nobody had added it up, because consultants’ work on proposals was not recorded anywhere as a cost.

An analysis of two years of losses showed three patterns: procurements in which the company was the third bidder invited as a benchmark; clients without an approved budget, collecting proposals in order to estimate it; and tenders in which the requirements were written so generally that an off-the-shelf solution formally met them. All three could be recognized quite quickly, but they were not treated as a problem, only as a market opportunity.

Qualification before proposing, not after. A modified MEDDICC implemented before the consultants sit down to work: who initiated the project and why now, whether the budget is approved, how many vendors were invited, who decides.

Written disqualification criteria. Four situations in which the company does not bid. Writing them down turned out to be more important than agreeing them, because it took the burden of justifying every refusal off the salesperson.

A diagnostic conversation instead of collecting requirements. Instead of accepting a list of functions, the company started establishing what specific problems the client is struggling with and what results they care about.

Two response formats. A full proposal with analysis only for qualified clients; for the rest, a short document prepared in an hour.

The company submits half as many proposals. Some enquiries end with a conversation after which both sides agree there is no point in going further, and that is treated as a good outcome. The position has changed too: a company that asks hard questions and is able to say no is perceived differently.

A conversation that did not exist before. The first contact after an enquiry comes in:

The proposal win rate rose from about 15% to about 45%, with the number of proposals submitted falling by roughly half. In total: more procurements won with significantly less work.

Several hundred hours of consultants’ work a year went back to projects that someone pays for.

Nearly forty people, an eight-year history, stable profitability. Two clients, both present for more than four years, together accounted for about 70% of revenue. The company was almost entirely tailored to the needs of those clients.

The owner knew that such a revenue structure was a risk, and regularly put the topic off, because serving those two clients took up all the available time. The turning point came when a friend’s company in a similar situation collapsed after one large client left. For years the company had been riding on stable contracts with key clients and generating profit, but the key clients knew perfectly well how important they were and used that fact at every opportunity.

The company had been stable and profitable for years, so there was no impulse to build sales. Smaller clients came from referrals and gave an illusory sense of market independence. The company had no marketing, did no active selling, and had no salespeople despite its already medium scale. The owner was often involved in coordinating delivery for the key clients.

Analysis of the company’s operational capabilities. What the company actually delivers, and for whom else it could be of value. Three elements of a value proposition emerged, but we focused on one.

Two new client segments. One close to the existing experience, the other clearly further away, deliberately, so as not to recreate the same concentration all over again.

Protecting the current portfolio in the meantime. We changed nothing in the cooperation with the current key clients; we built alongside.

Prospecting done by the owner, then delegated. To build and test the new process, the owner was first taken out of project involvement and focused mainly on building sales. Once prospecting started working, it was handed over to another person, and the owner tested the further stages of the process we had laid out for him.

The company has contracts in two new segments, smaller than the existing ones, but at a clearly higher margin, because dominant clients are not exploiting their position. Prospecting runs in a steady rhythm.

The share of the two largest clients fell from about 70% to about 25%, not by losing them, but by winning new clients.

A change that cannot be written in percentages: the owner stopped treating every phone call from the client’s IT director as the beginning of a crisis.

Calculating profitability separately revealed that part of the team’s work for one of the large clients never made it into the billing. That changed the way of reporting across the whole company. The second thing turned out to be the team structure: entering new segments required a model in which some competences are shared.

Twelve people, a good service product, loyal clients. Everything related to sales fitted in three places: the owner’s head, his mailbox and a notebook.

The problem surfaced with a simple question: how many enquiries have to come in to sign one order? There was no answer. Nor to how long the process takes, how many proposals the company loses and why. And the owner was considering hiring a salesperson.

We reconstructed nine months from the mailbox, the calendar, invoices and the notebook. Three things came out.

The company was losing opportunities not in the conversation, but after it: quite a few enquiries ended with a proposal followed by no contact from either side. Those were not losses, but opportunities that had been forgotten. Second, an enquiry from a regular client and from someone completely new came in through the same channel, and the existing client had priority of attention. Third, the company did not know why it was losing: the assumption was “because of price”, and in seven verified cases price was the reason in two. In the others it was the delivery lead time.

Process first, tool second. The company had once already attempted to implement a CRM, but after a few weeks it stopped being used and the old habits returned. Nobody had established how the process runs. We mapped five stages, and only then set up the system.

A minimal data scope. Four mandatory fields: stage, value, next step with a date, reason for closing. In a small company an overbuilt CRM at the start is a guarantee that nobody will use it.

A mandatory next step with a date. This one change accounts for most of the recovered opportunities. The system forced the next action to be defined, and it then landed on the to-do list.

A verified reason for each loss. Not guesses, but a short call to the client with one question. Almost everyone answers.

The owner sees the whole state of sales on one screen and does not have to remember anything. Opportunities do not get lost, because each one has a planned next step. The company knows its numbers and its conversion rate, so the plan for the next period stopped being a wish and became predictable. It also knows the real reasons for losses: the most frequent turned out to be the delivery lead time, not price.

About a fifth of the opportunities that previously vanished came back into the process. Some ended in an order, with no new acquisition activity, purely because someone went back to them. On an annual basis that is close to a million zlotys of additional revenue, with no additional acquisition costs.

The company is actively recruiting new people for delivery, and the seed of a sales team is forming.