An empty boardroom in early morning light.
Spine resolves

We resolve the organisation design questions that carry real consequence.

Scroll
We are

A specialist organisation design firm based in Europe. We work on teams, functions and divisions of a few hundred to a few thousand people, in mid-sized and large multinational organisations, and for their private equity owners.

We help them decide on

Operating models, structures, spans and layers, sizing, roles and responsibilities, and who decides what.
And once AI agents do part of the work, who checks what they produce and who is accountable for it.

Our services

What organisation design question is on your mind?

We help you answer one of these questions in a few weeks, without starting a large transformation programme. That is possible because our own OD AI tools and methods accelerate the data analyses, so a senior organisation design expert can focus completely on applying judgement to what is right for your company.

More information about our services →

An example

An example: a consumer food brand of 450 people with AI agents in its commercial team

THE QUESTION WHAT THE LEADERSHIP TEAM ASKED We put AI agents into the commercial function last year. The work changed,the organisation did not. Nobody can tell me what it should now look like,or who is answerable when an agent writes the bid a customer signs.THE AMBITIONA commercial organisation sized for the work that is left,with one named owner for every capability and every decision,and a leadership team able to approve it in a single session.01One workflow, all the way downchosen on hours and exposure02The function around itcapabilities, layers, spans, rights03One recommendationand the case against it HOW THE FOUR WEEKS WENT Four weeks from the first call to the decision paper, for a fixed price of €35,000 agreed before we started, and without workshops.
WHERE THE WORK SITS 84 people · 6 workflows · every one sized SHARE OF THE FUNCTION’S HOURSOF THAT, WHAT AN AI AGENT CAN DRAFTQuote to contract41%62%taken all the way downRenewals and re-pricing17%55%Account planning14%20%Campaigns and content11%70%Forecasting and reporting9%80%Onboarding handover8%35% what an AI agent can draft what still needs a person All 6 were sized. One was taken all the way down to steps, owners and signatures: the most hours behind it, and the only one that ends in a commitment.
TODAY 23 working days · 7 teams touch it · 6 handoffs EnquiryMarketingday 2QualifyInside salesday 5ScopeField salesday 9PriceBid deskday 13Write bidBid deskday 17Margin checkFinanceday 20ApproveCommercial dirday 22ContractLegalday 23 PROPOSED 7 working days · 2 teams · 1 handoff · 3 steps drafted by an AI agent Qualifyinside salesday 1Read RFP, match bidsAI agentday 1Scope the workbid managerday 2First-pass priceAI agentday 3Set price and margindeal desk leadday 4Draft the bidAI agentday 5Review, sign, senddeal desk leadday 7 produced by an AI agent, in minutes judged by a person, and one of them signs The 3 AI steps replace work that took 4 people 11 of the 23 days. The question the design answers is not what an AI agent can draft, it is who is answerable when it does.
WHO DOES WHAT MarketingInside salesField salesDeal deskFinance & legalAI draftsDemand generationownsQualifying an enquiryownsScoping the workownsPricing a dealownsProducing the bidownsContractingownsRenewalsowns does it today owns it in the design an AI agent drafts the first version 4 of the 7 happen in more than one place today, pricing in 3. In the design each has one owner, and 5 of the 7 now begin with something an AI agent produced.
TODAYPROPOSEDCommercial director84 peopleMarketing9 peopleInside sales14 peopleField sales38 peopleBid desk11 peopleNorth10South12Key accts145 layers to a salesperson · average span 3.4 · 8 managersCommercial director74 peopleMarketing9 peopleRevenue operations25 peopleField sales34 peopleDeal desk9Inside sales16North16South184 layers · average span 5.6 · 5 managers WHYQualifying happened in 2 places and pricing in 3, so every bid waited on handovers between them. One revenue operations team ends both, and that is where 16 of the 23 days sit. 84 roles to 74: 3 of the 8 manager posts, 4 in field sales, 3 in finance support. Key accounts folds into the 2 regions as named accounts.
WHO DECIDES WHAT WHAT MOVED Field salesDeal deskCommercial dirFinanceDiscount up to 15%RDIIwas the commercial director, 4 daysBid below target marginRRDCunchangedNon-standard payment termsIRCDwas legal, no service levelDeclining a bidRDIIwas nobody, bids driftedSigning a bid an AI agent wroteIDIInew, and the reason for the restPrice list changeCRCDunchangedD decides · R recommends · C consulted · I informed. 3 of the 6 moved, and those 3 are where 16 of the 23 days went.
THE RECOMMENDATION Approved in one session WHAT WE RECOMMENDEDMove pricing and margin authority into the desk that writes the bid, fold the bid desk and inside sales into one revenueoperations team, and put one named person behind every bid an AI agent drafts. One decision for the leadership team, not 5.23 → 7working days to a signedcontract, same approvals84 → 74roles, all of them managementor handoff work8 → 5managers, average span3.4 to 5.63 of 6decisions move, and 16 ofthe 23 days with them€800,000a year of payroll, against aone-off fee of €35,000 THE CASE AGAINST IT, WRITTEN BY US The revenue operations lead is a new job carrying pricing authority, and there is no obvious internal candidate. If this post is not filled properly the 7 days of gain are not realised. Getting there is not free. Severance on 10 roles runs to roughly €700K, so most of the first year pays for the change and the payroll only runs clean from the second.

Here are more examples →

Insights

What we have written

6 min read

AI agents change the work. The organisation stays as it was

When an AI programme fails, the cause is often the organisation around it rather than the tools. The structure around them was designed for a company where people did every step.

Read it →

4 min read

Best of both is the most expensive way to integrate two companies

A synergy case assumes the combined organisation is the better half of each. In practice it is a third operating model that neither company has run before, and learning to run it can cost about as much as the synergy was supposed to save.

Read it →

4 min read

In retail, head office cost is not a head office problem

Support functions get cut one at a time, against a benchmark percentage of sales. The cost they actually create sits in the store manager’s week, and nobody owns that.

Read it →

3 min read

A manager with 6 people and 12 AI agents does not have a span of 18

Span of control was a proxy for how much supervision one person can carry. AI agents make that count unreliable, and most companies still count span the old way.

Read it →

More articles →

Contact

Send us your organisation design question

Send us a short email on the decision you face and by when. We reply within one working day, and after one call you get a fixed price and a date, or the name of a firm that fits better.

info@spineresolves.com