A target process for launching educational programmes

Thesis Partners designed a target process for launching new educational programmes for Central University: decision points, a clear split of responsibility between functions and a shared launch calendar.

The client

Central University is the first university in Russia built on the STEM model, with a campus in Moscow and bachelor’s, master’s and executive programmes. Its tracks — artificial intelligence, software development, business and analytics, design — are led by working practitioners. The university was founded with the participation of technology companies, and people who know the market’s requirements from the inside take part in designing its programmes.

The results are visible beyond the classroom: the university’s researchers publish applied studies, from computational methods in oncology to computer-vision systems developed together with Skoltech, and the student and school teams trained at the university win awards at international olympiads.

At the same time the university is young: it has been admitting students since 2023 and is growing fast — in programmes, tracks and staff. A large part of the team combines teaching with industry practice.

The task

Much of the university’s growth rests on its culture: a high level of trust between teams, a readiness to help and to take responsibility, short distances between people. That culture is an asset, and the client’s request was framed in exactly that logic: to fix the way new programmes are launched so that it supports the culture rather than replacing it with bureaucracy.

As an organisation grows, the task naturally gets harder. Once there are more participants than personal acquaintance can hold — in social anthropology that threshold is known as Dunbar’s number — coordination starts to need support beyond personal agreements. The university came with its request in advance, before that became a constraint.

In substance, the work required: defining the moments at which a launch decision is taken and who takes part; separating responsibility for the launch route from responsibility for the programme’s content; bringing the functions’ plans into one calendar; and assembling a minimal set of working documents. All on condition that the process stays light.

The approach

The work took two months and ran in three stages.

Diagnostics. More than twenty interviews with the heads of twelve functional areas: product, academic leadership, finance, legal and methodology, production of educational content, faculty recruitment, brand, marketing, admissions and enrolment, the study office, partnerships and programme operations. The result was a factual picture of a launch: who joins at what moment and what inputs each function needs.

Design. The target model was assembled iteratively together with the university’s teams: every version was reviewed with the functional heads, comments were incorporated and taken to the next round. There were several full iterations with the working group and two discussions at the steering committee.

Documentation and handover. The process was fixed as a map on three A2 sheets, a normative description and a set of working templates. An adoption model was proposed separately.

The result

Four decision points on the annual launch cycle: concept, economics, communication, readiness to start. Each has a defined input, a set of participants and possible outcomes; the decision is recorded in a single place. The model is deliberately light — four decisions instead of multi-step approvals.

A separation of roles in the launch pair. The owner of the route is the product manager: calendar, working group, status, escalations. The academic lead is responsible for the programme’s content. Each focuses on their own area, and for the remaining functions responsibility became readable.

A shared launch calendar counting back from the start date, with a traffic light for deadlines and dependencies: the functions’ plans, which used to live in different horizons and formats, are brought onto one timeline.

A set of working tools — six documents: the launch procedure, the programme concept, the programme passport, a financial mini-model, a single launch brief and the calendar. Assembled by the rule “nothing is written twice”.

A process map on three A2 sheets: the main annual route, a short route for launching a cohort or a format on an existing programme, and a partnership module.

The value of a decision point is not in control but in the ability to change or postpone a launch while that is still cheap. That is why all outcomes in the model are equal: proceed, send back for revision, postpone, close — and each is recorded with a reason. This is what distinguishes a light process from bureaucracy: there are few decisions, but they are real.

Nikita Nechaev, partner at Thesis Partners

Adoption

The project ended with the handover of the model and a proposal for its adoption — that is separate work with a different logic, and we described it separately.

The proposed sequence: the next launches run on the new model; three to five internal champions from among the product managers and academic leads learn the route first and help their colleagues; regular oversight is assigned to the chief operating officer; after the first launches a retrospective is held, and the process owner inside the university updates the rules and templates on its basis. The check runs on two real launches.

An organisation adopts a process model when every function sees its own benefit in it: finance gets an early signal about the workload, legal gets the chance to check constraints in advance, admissions gets the enrolment parameters at the start of the design work. That is why we spent most of the time not on the diagram but on agreeing the wording with the functional areas: a process the team cannot read in its own words does not work.

Boris Kaptelov, partner at Thesis Partners

We grew fast and wanted to lock in what already works — trust between teams and the ability to reach agreement quickly. The value of the project for us is that the process came out short: a few clear decisions, a shared calendar and clear roles. It frees up time for the content of the programmes instead of adding approvals.

Alexander Ponomarenko, Central University

The effect

Launching a programme became transparent: at any moment it is clear what stage each programme is at, who owns the route and what decision has been taken. Decisions became timely: questions of budget, resources and public commitments are discussed at the moment when changes are still cheap.

The functions’ work is synchronised: all launch participants share one timeline and one source of status instead of parallel versions. The experience became reproducible: each next launch starts not from reconstructing the context but from a ready structure — and that speeds up the onboarding of new staff into project work.

Scale of the project

Two months of work, more than twenty interviews, twelve functional areas. The output: four decision points, six working documents and two pilot launches to test the model.