7) Stine Haugseth: Insourcing, retning og strategi på en ølbrikke

Truls Jørgensen og Audun Fauchald Strand — Heis Podcast · podcast · 2025-12-15 source ↗ Reader 🖥️

Data Sovereignty Go to the Floor (Gemba)Insourcing & In-House Capability Platform Engineering Professional Identity Public-Sector Digitalization Strategy & Direction-Setting Team Autonomy

Insourcing, retning, and strategi-på-en-ølbrikke are three faces of one leadership move: pulling the locus of decision-making in-house and then compressing direction tightly enough that a distributed organisation can actually act on it — insourcing repatriates the people who must own the work, oppdragsbasert ledelse and sirkler av autonomi give those people a clear mission with explicit frihetsgrenser, and the ølbrikke forces utviklingsstrategi down to the few hovedområder a development organisation can actually carry, with OKRs underneath and an explicit business case underneath that.

practical · narrative-expository — a leadership case interview built around three intertwined arcs: an autobiographical arc (NAV-analytics → forsikring → Spink revelation → Mattilsynet rebuild → SpareBank 1 Østlandet), an organisational practice arc (the operational mechanics of insourcing, oppdragsbasert ledelse, sponsor-vs-styringsgruppe, sirkler av autonomi), and a compression-as-method arc (utviklingsstrategi reduced to an ølbrikke, plus OKR underneath). Hosts ask short questions; Haugseth answers at length. Practical-leaning throughout; argumentative on insourcing's business-case requirement and the sky/skye reversal; narrative on the Mattilsynet contract cancellation window, the Spink team, and the Sporveien driftssjef. — Both hosts have shared NAV history with Haugseth ('jeg jobbet i Ønsenjongen der du jobber i dag... tror jeg nesten samme avdeling'), which surfaces in shorthand references to NAV's 2016 insourcing journey and to NAIS-as-a-service.

Structure

Part	Role
Opening: CV (NAV analytics → forsikring → Mattilsynet → SpareBank 1 Østlandet) and the entry question 'hvordan kan ledere finne den vanskelige balansen mellom frihet og styring, visjon og drift, menneske og struktur.'	Frames the entire conversation around the frihet/styring tension that insourcing, oppdragsbasert ledelse, and the ølbrikke all try to resolve.
Lederstil evolution: from 'sort-hvit, dette er min ene lederstil' to a repertoire chosen by team/situation; favoritten er den visionære (give technologists the freeway and clear obstacles); det verste er when someone is not functioning in their role.	Establishes that her later prescriptions about teams, autonomy, and visjon come from someone who explicitly diagnoses what stil a situation needs.
Public vs. private sector: 'det er jo folk' — the same people. What differs is the utløsende hendelse for transformation (Mattilsynet's KPMG report; the previous administrerende having to leave).	De-mystifies the offentlig/privat divide and reframes it as a question of mandate width.
The Mattilsynet insourcing mechanics: large consultant contracts cancellable twice a year; took the window with only 3-4 months runway; first developer hired in ~2 months; scaled to 30-35 developers in 3 years. Cross-reference to NAV's 2016 insourcing being much slower at the start (1, 2, 3, 5).	Concretises insourcing as an operational move with timing, contracts, and pace — not a slogan.
Insourcing extended beyond developers: in Mattilsynet they also brought jurist competence in alongside developers, eliminating a regelverks-loop between organisations.	Generalises insourcing from a tech-only intervention to a colocation intervention.
SpareBank 1 Østlandet contrast: here she has NOT cancelled all consultant contracts; the bank relies on SpareBank 1 Utvikling for shared product development; only TWO Østlandet-specific product teams complement, focused on differentiating Østlandet (kundeutbytte; rådgiver-hverdagen).	Calibrates the insourcing prescription against organisational context — insourcing isn't a uniform answer.
Business-case requirement: insourcing must produce a measurable gevinst — either consultants get redeployed elsewhere, or rådgiver-produktivitet boosts inntekt. 'Du kan ikke bare gjøre det for å gjøre det.'	Imposes accountability on the insourcing move; rules out insourcing-as-ideology.
Oppdragsbasert ledelse — from a Forsvars-foredrag many years ago; Nordic vs. old Soviet military model; clear mål let styrkene improvise underveis; trinn-for-trinn collapses under reality. Forsterket av Ukraina/Russland-erfaringene.	Provides the conceptual lineage for how she does retning with technology teams.
Concrete oppdragsbasert mechanic: a one-pager oppdragsbeskrivelse signed by KL/toppledergruppe per team, plus a small sponsorgruppe (2 dedicated toppledere with medansvar) instead of a heavyweight styringsgruppe. Team-eierskap shifts from 'Stine sine team' to 'våre team.'	Names the actual artifact and forum that makes top-level retning concrete and shared.
Scaling problem: this requires a lot of toppleder-presence; doesn't scale to 100 teams. Hence produktområder (her NAV term) as an intermediate framing layer.	Acknowledges the limit of the one-pager+sponsor model and points at a structural answer.
Sirkler av autonomi (Netflix reference, pre-'aligned autonomy'): team decides everything inside the circle; cross-area things require coordination; only a few things go to top level. 'Du må prøve å legge så stor frihetsgrad som mulig til teamet.'	Generalises the team-rammeverk move beyond the per-team one-pager.
Ledergruppens rolle i smidig: warning that a teknologi-ledergruppe can drift into a 'sånn administrativ enhet'; the avdelingsledere/seksjonsledere must be ansvarliggjort for cross-team samhandling. 'Det innser jeg nok litt sånn sent i Mattilsynet.'	Self-critical observation about a failure mode in her own past — the parallel-structure problem.
Spink (SpareBank 1's smart car insurance, now discontinued): a tiny cross-functional team of 3-4 producing 'enormt mye' compared to her then-22-person team; revelation about ownership and tverrfaglighet shapes how she builds teams afterward.	Origin story for her team-design instincts — small kjerne, draw in resources as needed.
Career advice digression: tør å mene noe, observe leaders (also the negative examples), take the opportunity actively. No one offers leadership; you have to seek it.	Brief expository aside; supports the autobiographical thread.
Skye-reversal: she used to be 'veldig pådrevet' for cloud (AWS/GCP); now more critical — beredskap, fallback, geopolitikk, kostnad-lock-in. Finanssektoren is 'heldige' to be lovpålagt exit-strategi. Implicit call: every tech leader should ask 'hvor godt gifte er jeg med en storteknolog?'	Concrete example of changing one's mind based on Mattilsynet beredskap-work and recent kostnadshits.
NAIS-as-a-service reflection: she was 'veldig ivrig' that Mattilsynet pilot NAV's platform; the decision went another way; in retrospect she would have gone with NAIS sooner. The most valuable thing was not the technical platform but the community / grupperingen around it.	A concrete forangerelse: an admitted strategic miss and a counter-prescription for offentlig sektor (piggyback on the large players' platforms rather than building parallel plattformer).
Visjonær lederstil as method (not innate): you must ask 'hvorfor skal jeg lede dette?' and 'hvorfor skal vi på jobb?' Inspirations: Torbjørn Larsen at NAV; a Sporveien T-bane driftssjef who drilled 'vi kjører T-banen' into every newcomer.	Operationalises visjon as a repeated, embodied alignment ritual, not a powerpoint.
The ølbrikke method: an utviklingsstrategi compressed to a beer coaster — three hovedområder (eierskap, egen art, lønnsomhet) with mer konkrete strategiske føringer below, OKR-rammeverket as the operationalisation layer.	Names the central method of the episode; ties visjonær compression to a concrete artifact.
The balancing question: include too many in strategy-making and it goes lyserådt; do it alone and it doesn't anchor. Her answer: overordnet visjonær on the toplevel, more concrete strategiske føringer below, OKRs underneath.	Addresses the classic strategy-process problem and ties it back to the abstraction-altitude principle.
Closing: lederrollen som privilegium, not identitet; 'ikke få tett og klamt eierskap til en lederrolle.' At a point, someone else may be better suited for a steady-state phase.	Personal coda; reinforces the role-shaped-by-situation theme from the opening.

Key terms

Insourcing (hjemtaking) Insourcing & In-House Capability
Repatriating the locus of work — and the decision-making that comes with it — by terminating consultant contracts and hiring in-house. In Haugseth's operational vocabulary, insourcing is governed by (1) a window (when can you cancel? Mattilsynet's 2-per-year cancellation gates), (2) a snowball effect ('hvis du får tak i noen skikkelig flinke folk... så vil de kunne få tak i enda flere'), (3) a competence prerequisite ('du må vite hvem du skal insource' — kvoting Torbjørn Larsen: 'du må ha peiling for å skrive på peiling. Også for å ansette peiling'), and (4) a business case ('konsulentene får ny oppdrag andre steder, ELLER vi får boostet inntektene våre').
Retning Strategy & Direction-Setting
Direction-setting in a way that is intelligible at the team level — concrete enough to act on, abstract enough to leave room for execution choices. For Haugseth, retning is enacted through (a) oppdragsbeskrivelser, (b) frihetsgrenser, (c) compressed strategy artifacts like the ølbrikke, and (d) ritual repetition (the Sporveien T-bane example).
Strategi på en ølbrikke Strategy & Direction-Setting
An utviklingsstrategi (not virksomhetsstrategi) compressed to fit on a beer coaster — three or so hovedområder named explicitly (eierskap, egen art, lønnsomhet in Mattilsynet's case), with more concrete strategiske føringer below and OKRs as the operationalisation layer. The ølbrikke is a method, not a metaphor: they literally printed it. Its function is to make the development-side translation of the corporate strategy short enough that everyone can carry it.
Oppdragsbasert ledelse Team Autonomy
Mission-based leadership borrowed from the (Nordic) military: leaders specify mål and rammer; subordinates choose how to reach them. Haugseth contrasts it with the trinn-for-trinn (Soviet) model that collapses under contact with reality, and cites Ukraine/Russia as recent reinforcement. Operationally: a one-pager oppdragsbeskrivelse from konsernledelse to each team, with a small sponsorgruppe — NOT a styringsgruppe.
Sirkler av autonomi Team Autonomy
Concentric authority structure: the team decides everything inside its circle; cross-team things are negotiated horizontally; only a few high-altitude things require top-level decision. Haugseth attributes the framing to a Netflix presentation she saw 'før vi begynte å snakke om aligned autonomy.'
Sponsorgruppe (ikke styringsgruppe)
A small group (two members) from the toppledergruppe explicitly co-responsible for a team's success — sitting beside the team rather than above it. Designed to spread eierskap across konsernledelsen so that the teams stop being 'Stine sine team' and become 'våre team.'
Hjemtaking av jurister Insourcing & In-House Capability
An under-emphasised aspect of Mattilsynet insourcing: jurists were brought in alongside developers so regelverksutvikling stopped being an external loop. The principle generalises: insourcing's value is partly about which inter-organisational loops it collapses.
Visjonær lederstil
Her favourite style: give technologists the freeway, clear obstacles, hold the ambisjon high enough that they can run. Method (not innate): repeated articulation of 'hvorfor er vi på jobb' until everyone carries the same answer — the Sporveien driftssjef's 'vi kjører T-banen' as the archetype.
Hospitering Go to the Floor (Gemba)
Sitting in teamrom, walking the gulvet, listening — her main learning loop as toppleder. Yields arbeidsglede, impulser, and exposes her to genuine teknolog-disagreements that re-shape her world model.
Skye / sky-strategi Data Sovereignty
Her changed view: from full pro-AWS/GCP advocate ('bare kjører vi') to 'kanskje ikke alt skal opp i skye' — driven by beredskap-arbeidet at Mattilsynet, geopolitisk endring, and kostnadshits. The corrective: exit-strategi, beredskap, fallback-løsninger. Finanssektor is lovpålagt to have exit-strategi; she wants everyone to think about it.
NAIS as a service Platform Engineering
NAV's application platform (NAIS), offered as a tjeneste to other public etater. Haugseth advocated piloting it at Mattilsynet; the decision went another way. In hindsight: she would have used it. The most valuable thing is not the plattform itself but the community / grupperingen — a way of thinking about verdiforbruker and forvaltbar kode over tid.
Business case for insourcing Insourcing & In-House Capability
Insourcing must show measurable gevinst — either consultants are redeployed productively elsewhere, or rådgiver-hverdagen improves enough to boost inntekt. 'Du må vite hva du skal ha som mål med det.'
Privilegium, ikke identitet Professional Identity
The toppleder-rolle is borrowed time. 'Ikke få tett og klamt eierskap til en lederrolle.' Someone else may be better for a steady-state phase later.

Propositions

explicitInsourcing is the right move for software competence specifically — you save money AND you couple the developers tightly to the organisation.conf: high
explicitInsourcing's window depends on contract structure — when consultant contracts allow cancellation only twice a year, leaders must seize the window even with thin runway.conf: high
explicitInsourcing snowballs once it has started — flinke folk attract flinke folk, so the early hires are disproportionately load-bearing.conf: high
explicitInsourcing should not be done for its own sake — there must be an underliggende business case (redeployment of consultants, productivity gain that boosts revenue).conf: high
explicitInsourcing should extend beyond developers — colocating jurists with developers collapses inter-organisational loops and is itself an insourcing move.conf: high
explicitThe right insourcing depth depends on context — SpareBank 1 Østlandet keeps most software at the shared SpareBank 1 Utvikling and only insources for what is genuinely Østlandet-differentiating.conf: high
explicitPublic-sector insourcing is no harder than private-sector insourcing — what differs is the utløsende hendelse (e.g., a KPMG-rapport, an administrerende leaving) that creates mandate width.conf: high
explicitOppdragsbasert ledelse — give clear mål and rammer, not trinn-for-trinn — is the operating model that survives contact with reality; the Nordic military model has outperformed the old Soviet model precisely on this dimension.conf: high
explicitA one-pager oppdragsbeskrivelse from konsernledelse, plus a two-person sponsorgruppe (not a styringsgruppe), is a smart institutional vehicle for shared eierskap of cross-cutting teams.conf: high
explicitThe one-pager + sponsor model requires high toppleder-presence and does not scale to 100 teams; an intermediate produktområde layer is needed at scale.conf: high
explicitA teknologi-ledergruppe risks degenerating into an administrativ enhet unless avdelingsledere/seksjonsledere are explicitly ansvarliggjort for cross-team samhandling.conf: high
explicitSmall, tverrfaglige, owner-of-product teams dramatically outperform larger functional teams — the Spink team of 3-4 produced 'enormt mye' compared to Haugseth's then-22-person team with ~10x the budget.conf: high
explicitVisjonær leadership is not innate — it is a learnable ritual of repeating 'hvorfor er vi på jobb?' until every newcomer carries the same answer; the Sporveien T-bane driftssjef who drilled 'vi kjører T-banen' into every newcomer is the archetype.conf: high
explicitUtviklingsstrategi should be compressed to fit on an ølbrikke — three hovedområder (eierskap, egen art, lønnsomhet) with concrete strategiske føringer below and OKRs as the operationalisation layer.conf: high
explicitIncluding too many in strategy work makes it lyserådt; doing it alone leaves it unanchored — the balance is overordnet visjonær at the top, more concrete strategiske føringer below, with OKRs underneath.conf: high
explicitTech leaders must understand technology well enough to recruit and to insource — without that competence, insourcing produces a 'veldig rart miljø' that ends up dependent on external consultants again.conf: high
explicitdevelopedSkye is no longer the obvious answer — beredskap, geopolitisk lock-in, and kostnadshits all argue for fallback-løsninger and exit-strategi; finansbransjen is lucky to be lovpålagt this discipline.conf: high
extends ←Forsvaret med egen data-grunnlov · p3 — Haugseth named beredskap and geopolitisk lock-in as reasons finance is moving toward fallback-løsninger/exit-strategi from cloud vendors; Forsvaret's FDSA operationalizes exactly that same logic as formal, named architecture, and Minister Tung explicitly proposes generalizing it beyond one regulated sector to the whole public sector.
explicitSmaller offentlige etater should piggyback on the large players' platforms (NAIS) rather than build their own — the most valuable thing they get is not the plattform but the community / grupperingen around it.conf: high
explicitcorroboratedHospitering — sitting in teamrom, walking the gulvet — is how a toppleder stays epistemically grounded; disagreement among technologists is the highest-leverage learning input.conf: high
supports ←What New CEOs Get Wrong About the Job · p4 — Nohria's 'go to the front lines, talk to customers' advice is the English-language, CEO-specific instance of the same gemba/hospitering discipline Haugseth describes as how a toppleder stays epistemically grounded.
explicitLederrollen is a privilegium, not an identitet — leaders should resist 'tett og klamt eierskap' to the role, recognising that someone else may be better suited for a different phase.conf: high
implicitDirection-from-the-top in a smidig context is itself the team's frihetsgrenser — naming what is inside the circle, what coordinates horizontally, and what escalates up.conf: high
implicitA leader's repertoire of stiler is a deselection problem in itself — knowing which style not to use for a given team/situation is part of the craft.conf: medium

Arguments

Insourcing is the right operational move for software-heavy organisations — but it must be governed by a business case, a competence prerequisite, and an exploitation of the contract window. strength: high

Explicit premises:

Software competence pays back faster than other competence categories when brought in-house.
Insourcing requires the leader to know who to hire — i.e., the leader needs enough tech understanding to filter candidates.
Insourcing must produce a gevinst — consultants redeployed elsewhere, or productivity boosting inntekt.
Contract structure dictates the timing — Mattilsynet had only twice-yearly cancellation windows, which forced the move on a 3-4 month runway.

Implicit premises:

The leader's epistemic warrant for insourcing decisions is part of the move itself; an insourcing leader without tech literacy reproduces consultant-dependence at a new layer.
Insourcing's snowball effect — flinke folk attract flinke folk — makes the first hires disproportionately load-bearing.
Insourcing creates the conditions under which oppdragsbasert ledelse and small cross-functional teams become viable.

Inferences:

A leader without a business case for insourcing should not insource.
A leader without tech literacy should not be the one driving insourcing decisions.
Insourcing's value depends partly on which inter-organisational loops it collapses (e.g., bringing jurists in alongside developers).

Retning compressed to a one-pager and an ølbrikke does more for direction-coherence than heavyweight strategy documents or trinn-for-trinn instructions. strength: high

Explicit premises:

Trinn-for-trinn collapses under contact with reality (military example).
An oppdragsbeskrivelse + small sponsor group gives the team a clear mission with concrete rammer.
An utviklingsstrategi on an ølbrikke compresses direction to what people can carry; mer konkrete føringer and OKRs do the operationalisation below.

Implicit premises:

Compression is a forcing function for deselection — what doesn't fit on the ølbrikke isn't really strategi.
Direction held only by senior leaders is direction the gulvet cannot act on.
Repetition is the medium by which compressed direction becomes shared (the Sporveien example).

Inferences:

If your utviklingsstrategi cannot fit on a coaster, you have not yet done the strategy work.
If your strategy does not propagate down to OKR-level work, you have a powerpoint, not a strategi.
The combined oppdragsbeskrivelse + ølbrikke + OKR stack is itself a way of distributing decision-rights at different altitudes.

Sponsor groups (not styringsgrupper) and shared eierskap across konsernledelse are the institutional vehicles by which 'Stine sine team' become 'våre team' — without which transformasjons-arbeid runs as a parallel structure that the rest of the organisation ignores. strength: high

Explicit premises:

Smidig transformasjon easily creates a parallel struktur the rest of the organisation lets live fritt.
A two-person sponsorgruppe with medansvar invests konsernledelsen in the team's success.
Avdelingsledere/seksjonsledere must be ansvarliggjort for cross-team samhandling, not allowed to drift into administrasjon.

Implicit premises:

Without shared eierskap, the transformation lead becomes the only person accountable, which is unstable.
Ledergrupper that are not held responsible for cross-team samhandling will degenerate into status-meetings.
Shared eierskap is itself a leadership artefact — it requires deliberate design (the sponsor model) rather than just expectation.

Inferences:

If your smidig transformasjon is led by a single person's team, it is institutionally fragile.
If your ledergruppe meetings don't deal with cross-team samhandling, your operating model has a structural gap.
The right size of a 'sponsor' is small — a heavyweight styringsgruppe re-imports the parallel-structure problem.

Insourcing logic extends to the platform layer — sky-strategi is itself a question of where the locus of control should live. strength: medium

Explicit premises:

Geopolitical and cost dynamics now argue against full sky-avhengighet.
Beredskap, fallback, and exit-strategi are increasingly required.
Finansbransjen is lovpålagt exit-strategi; other sectors should adopt the discipline anyway.

Implicit premises:

The same logic that justifies insourcing software competence — keep the locus of decision-making close — applies to infrastructure.
A leader's view of where 'verden beveger seg' (cf. Olrud Moen ep 15) must include geopolitical and supplier-concentration trends, not only product/market trends.
Sky-decisions made on cost-per-compute alone underweight the cost of lock-in.

Inferences:

If you cannot articulate your exit-strategi from your hovedskyleverandør, you are over-leveraged.
If your sky-strategi has not been re-examined since 2021-2022, it may be running on premises that have shifted.
Sovereign / European / Nordic cloud alternatives deserve active utforskning.

The insourcing prescription has a limit — small offentlige etater should NOT replicate the plattform-bygging of the large ones; the highest-value piggyback is community, not code. strength: medium

Explicit premises:

NAIS at NAV represents enormous investment that smaller etater cannot replicate well.
The most valuable thing NAIS-as-a-service brings is not the technical platform but the surrounding community / grupperingen and its way of thinking about verdiforbruker and forvaltbar kode.
Mattilsynet's decision to not pilot NAIS is, in retrospect, one Haugseth would have made differently.

Implicit premises:

Communities of practice transfer tacit knowledge that platforms alone cannot encode.
The 'build vs. buy' decision underweights 'join a community' as a third option.
Insourcing as a frame can mislead by treating ownership-of-stack as the only locus question.

Inferences:

A small etat that builds its own plattform pays twice — once for the platform, again for the missing community.
Platform-as-a-service offerings should be evaluated partly on the strength of their surrounding community.
Haugseth's general insourcing pro-stance has a clear, named exception.

Lederstil should be chosen for situation, not signature — visjonær for high-ambition technologist teams, more direktiv for those who need clearer rammer, harder/stricter for handling underperformance and the 'kjipe' part of the role. strength: medium

Explicit premises:

Different mennesker, teams and organisations respond to different lederstiler.
Haugseth's favoritt is visjonær; her least favourite but necessary work is handling people 'i russ.'
Diagnosing which stil a situation needs is itself part of the leadership craft.

Implicit premises:

A signature lederstil is a refusal to do the diagnostic work the situation requires.
The role is borrowed — even one's preferred style should not become identity (the 'privilegium ikke identitet' frame).
Treating leadership as a repertoire is what makes oppdragsbasert ledelse possible — the same leader can move between altitudes.

Inferences:

If you only have one lederstil, you are limited to the situations it fits.
If you cannot do the kjipe oppgavene with the same craft as the visjonære ones, you cannot hold the role.
The right time to leave a role may arrive before you stop being effective in it.

Narratives

Mattilsynet insourcing: Haugseth arrives roughly a year after Inger Helene (predecessor's predecessor — name uncertain in transcript) was tasked with cleaning up Mattilsynet following a graverende KPMG-rapport that cost the previous administrerende the role. Endringsløpet is 'opp for grabs'; the team identifies that the large consultant contracts are cancellable only twice yearly. They take a window with 3-4 months runway, hire the first developer within two months, and scale to 30-35 developers over three years — alongside the parallel insourcing of jurist-kompetanse into the developer teams.
→ Concrete operational case for the insourcing prescription — including the timing constraint, the snowball effect, and the generalisation beyond software-only.
NAV insourcing 2016: started slowly compared to Mattilsynet — 1, 2, 3, 5 — before scaling. Reference point for how unusual the Mattilsynet pace was.
→ Counterpoint and calibration — insourcing pace is not predictable; the right reference is your own contract structure and mandate, not someone else's curve.
SpareBank 1 Østlandet: insourcing is partial and bounded. The bank leans on the shared SpareBank 1 Utvikling for the bulk of product development; only TWO Østlandet-specific produktutviklingsteam exist, focused on differentiating Østlandet through kundeutbytte-løsninger and improving the rådgiver-hverdagen (the fragmenterte portefolio they navigate today).
→ Contextualises the insourcing prescription against organisational reality — the right depth depends on what is genuinely differentiating.
Forsvars-foredraget: many years ago, an officer (forsvarsfyr) explained why the Nordic military model outperformed older Soviet doctrine — clear mål let styrkene improvise their route, while trinn-for-trinn collapses when reality interrupts. Haugseth has heard the lesson reinforced through reporting on Ukraine/Russland.
→ Conceptual origin story for oppdragsbasert ledelse as her preferred operating model.
The Spink team: while running a 22-person product-team in forsikring with enormous cost-burden and expectations, Haugseth discovers a 3-4 person cross-functional team called Spink (SpareBank 1's smart car insurance) producing 'enormt mye' with a tenth of the development capacity. Investigates how they work, sees they own their product end-to-end with tight tverrfaglighet, and uses the observation to reshape her own team designs afterward.
→ Origin story for her preference for small cross-functional teams; concrete antecedent to the later oppdragsbeskrivelse + sponsor model.
The Sporveien T-bane driftssjef: many years ago, when Haugseth was working at (Bærums)værnsnøyen / Sporveien, the operations chief in the T-bane division ran 1-1 or classroom sessions with every newcomer. The signature was a single drilled question: 'hvorfor er du på jobb?' If you answered 'jeg skal sitte og skrive litt kode,' the response was 'nei, du kjører T-banen.' The result: 'du gikk inn døra... at her er det noe annet her, som ikke du ser alle andre steder.'
→ Archetypal illustration of visjonær leadership as ritual repetition — the Sporveien officer is Haugseth's working model for how to compress and propagate retning.
The ølbrikke: Haugseth's team in Mattilsynet drafted an utviklingsstrategi with three hovedområder (eierskap, egen art, lønnsomhet) compact enough to fit on a beer coaster — and then literally printed it on coasters. Below the three hovedområder ran more konkrete strategiske føringer, with OKRs as the operationalisation layer.
→ The episode's titular method — strategy compressed to its essential, then layered down through concrete føringer and OKRs.
NAIS-as-a-service: Haugseth advocated strongly that Mattilsynet pilot NAV's plattform-as-a-service (NAIS) — there were ulike utvikler-preferences for thin-plattform vs. Golden Path. The decision went another way (Mattilsynet stayed on its tynne, krevende plattform). In hindsight she would have moved to NAIS sooner — the most valuable thing was not the platform itself but the community and grupperingen connected to it.
→ Admitted strategic miss; sharp counterexample to the 'build your own platform' instinct in smaller etater.
The skye-reversal: Haugseth was previously full pro-AWS/GCP ('bare kjører vi') for utviklingsverktøy and cost; now is markedly more critical. Drivers: beredskaps-arbeidet at Mattilsynet, recent geopolitisk endring, sterke prishit fra tek-leverandørene, and the question 'vil vi bruke skattepengene våre på det?' Finanssektor's lovpålagte exit-strategi is, in her view, a discipline more sectors should adopt.
→ Concrete worked example of a leader updating her view, and an extension of the insourcing logic to the platform layer.

Questions addressed

✅ What does insourcing actually deliver, and why is it the right move for software-heavy organisations?
Cost savings + tight coupling of developers to the organisation; collapses inter-organisational loops; with a snowball effect on hiring quality. But governed by a business case and a competence prerequisite — and not always full insourcing (see SpareBank 1 Østlandet).
✅ When should a leader take an insourcing window, and what determines pace?
Contract structure determines the window (Mattilsynet's twice-yearly cancellation gates); pace depends on local conditions (Mattilsynet scaled fast; NAV 2016 was much slower at the start).
✅ Does insourcing extend beyond developers?
Yes — Mattilsynet's colocation of jurists with developers eliminated a regelverks-loop; the principle generalises to any cross-organisational loop where colocation reduces coordination cost.
✅ How do you set retning for a smidig organisation without micro-managing?
Oppdragsbasert ledelse — one-pager oppdragsbeskrivelse from konsernledelse, two-person sponsorgruppe co-responsible for team success, sirkler av autonomi making explicit what is decided inside the team vs. coordinated vs. escalated.
✅ How do you scale the one-pager + sponsor model?
An intermediate produktområde layer (her NAV term) acts as the framing structure when there are too many teams for direct toppleder-presence.
✅ How do you keep utviklingsstrategi from becoming nok et dokument i skuffen?
Compress to an ølbrikke — three hovedområder readable in one glance, mer konkrete strategiske føringer below, OKRs as the operationalisation layer.
✅ How does a tech leader stay grounded as the role moves them further from the gulvet?
Hospitering — sit in teamrom, walk the gulvet, listen for technologist disagreements as the highest-leverage learning input.
✅ How important is the leader's own tech literacy in insourcing-heavy organisations?
Critical — without it you cannot recruit well or insource well; insourcing without tech literacy reproduces consultant-dependence at a new layer ('du må ha peiling for å skrive på peiling').
✅ Has the case for cloud-first changed?
Yes — geopolitisk endring, prishit, beredskap-krav all argue for fallback-løsninger and exit-strategi. Finansbransjen is lucky to be lovpålagt this discipline; other sectors should adopt it anyway.
⭕ When does the insourcing instinct go wrong?
Implicitly named in the NAIS-reflection: small offentlige etater should NOT replicate large players' plattform-bygging — they should piggyback. But the criterion for when to insource a stack vs. join a community is not generalised.
⭕ What about insourcing efforts that fail outright?
Not explored. The conversation is structured around successful insourcing (Mattilsynet, NAV) and against the build-your-own-platform instinct, but contains no narrative of an insourcing that produced bad outcomes.
⭕ What do you owe the consultants who lose oppdrag when you insource?
Only briefly named in the business-case framing ('konsulentene får ny oppdrag andre steder'). Whether that absorption actually happens, and what leaders owe the people whose contracts they terminate, is not analysed.

Gaps

incompleteHaugseth's account is built on insourcing successes (Mattilsynet, NAV). There is no narrative of an insourcing that failed or backfired — no analysis of when insourcing produced worse outcomes than the status quo. The closest is the NAIS-reflection (where she would have stayed less ambitious about own platform), but that is presented as 'should have piggybacked,' not 'insourcing failed.' The asymmetry shapes what is visible: insourcing's hazards (talent risk, two-year ramp where the org loses access to specialists, capability concentration in a few key hires who then leave) are absent from the framing.
Read next: Susanne Justesen / John Cutler on the cost of capability concentration; the Norwegian NAV 'Optimizing for Fast Flow' article (already in the user's library) for a parallel account that emphasises the work-system rather than the ownership question; Capacity Planning literature on staffing as a long-cycle bet; failure-case writing on insourced platform teams that lost their flywheel (Stripe / Etsy public postmortems are useful counterpoints).
incompleteThe business-case for insourcing is asserted but not specified. 'Either konsulentene get ny oppdrag andre steder, OR rådgiver-hverdagen improves enough to boost inntekt' is a frame, not an analysis. What kind of measurement? Over what horizon? Who validates the case 24 months in? The hard version of the case — that some insourcing pays for itself only in scenarios that don't materialise — is not engaged.
Read next: Roger Martin on the strategy-stack from where-to-play to how-to-win; cost-of-delay literature; Dan North on staffing-as-bet.
incompleteOppdragsbasert ledelse is presented as obvious, with the Nordic-vs-Soviet military analogy doing most of the rhetorical work. But the conditions under which mission command works (shared training base, mutual trust, common doctrine) are exactly the conditions large public-sector and bank organisations often lack. The transcript skips over how an organisation acquires those preconditions — i.e., the part that is hard.
Read next: Stephen Bungay 'The Art of Action' (the foundational text for Auftragstaktik in management); David Snowden on what context-shaping does that direct order cannot; Vidar Moe's account (Heis ep 5) for an adjacent leadership-style discussion.
closed byPresentation: The Human Scalability Problem: Why Your Teams Don't Scale Like Your Code — Addresses how an organization acquires the preconditions mission command assumes: trust must be deliberately engineered via relationship-building before task work, repeated core messages, leader-modeled vulnerability, and rituals as persistent social infrastructure.
incompleteThe ølbrikke method is presented as a virtuous compression artifact, but the failure mode — that radical compression elides important trade-offs and lets every reader project their preferred reading onto the three hovedområder — is not addressed. Olrud Moen (ep 15) discusses the strategy-as-PowerPoint failure mode explicitly; Haugseth's ølbrikke could fall into a sibling failure mode (the strategy-as-slogan trap).
Read next: Olrud Moen Heis ep 15 (recently extracted) — direct counterpoint on strategy compression; John Cutler's TBM on 'making the trade-off visible'; Roger Martin on integrative thinking; Donald Sull on 'simple rules.'
incompleteThe sponsor-group mechanism is described in the abstract; what happens when the two sponsors disagree, or when a sponsor leaves konsernledelsen mid-oppdrag, or when sponsor priorities conflict with what the team's konsernledelse-wide one-pager actually says, is not analysed.
Read next: Marty Cagan / SVPG on product strategy governance; Cutler on accountability culture and DRIs (directly responsible individuals).
incompleteHaugseth's epistemic update on skye is offered as an instance of changing her mind, but the underlying mental model that drove her earlier full-pro-cloud stance is not interrogated. What other current convictions might be in the same epistemic position now? The 'how does a leader earn forankring on the world' question (same as the Olrud Moen ep 15 gap) is unaddressed.
Read next: Annie Duke 'Thinking in Bets' (already in user's library) on calibration and the cost of being skråsikker about probabilistic claims; Philip Tetlock on superforecasting; Donella Meadows on systemic leverage points.
incompletePower and political economy are visible at the edges (the consultant houses' motstand to Mattilsynet cancellations; the 'kjipe' work of handling underperformance) but never named as power. The political cost paid by the consultant houses, by the developers whose subkulturer were dismantled, and by the avdelingsledere who had to take on accountability for cross-team samhandling is not analysed.
Read next: Cutler 'Power and Space' (TBM 413, already in user's library); Pfeffer 'Power' on organisational politics; Hamel/Birkinshaw on managerial innovation and its political cost.
incompleteThe 'privilegium ikke identitet' frame is named but not connected back to the operating-model questions earlier in the conversation. A leader who treats the role as borrowed time presumably should design the operating model so it survives their departure — i.e., the ølbrikke and the sponsor-group should ideally outlast the leader. That implication is not drawn.
Read next: Cutler on operating systems as artefacts that survive their author; Olrud Moen ep 15 on 'strategy as the few sharp trade-offs that nobody else has the authority to set' — and the corollary question of what the leader leaves behind.

7) Stine Haugseth: Insourcing, retning og strategi på en ølbrikke

Structure

Key terms

Propositions

Arguments

Narratives

Questions addressed

Gaps

Related documents