Persona Gaps at Helio Retail
Helio Retail spends $6.2M a year across public cloud and SaaS. A four-engineer Cloud Center of Excellence was stood up 18 months ago and owns 'cloud cost'. A finance analyst builds a monthly spreadsheet from the billing export and emails it to department heads. Procurement runs SaaS renewals on a separate calendar and negotiates on list price. Each product team nominates an 'account owner' who is a director with no access to billing data and no technical staff reporting to them. Symptoms: 60% of SaaS renewals are signed without consumption data, engineering teams ignore the showback because it does not reconcile to their P&L, and the CoE is now the bottleneck for every cost question.
1. Map the personas present in this situation against who actually needs to act, and identify the two gaps that are causing the most damage.
Read the worked answer
Present and functioning: FinOps Practitioner (the CoE, but doing too much hands-on work), Finance (one analyst, producing output nobody trusts), Procurement (active, but blind to consumption).
Present in name only: Product Owner - the nominated directors have accountability without data, tooling or levers, which is the worst combination; Leadership/Executive - absent, since no one has set an expectation that cost is a product-team concern.
Absent: Engineering as a cost-owning persona. The four CoE engineers are acting as a proxy for fourteen teams that actually make the spend decisions. Every optimisation therefore queues behind four people.
The two most damaging gaps:
- Engineering ownership. The people who choose instance size, retention period and query pattern are not in the loop and see the showback as an accusation rather than a tool. No amount of reporting fixes this; it needs cost visible in the tools engineers already use and an explicit expectation from leadership that they act on it.
- Procurement without consumption data. Signing 60% of renewals on list price is a measurable, avoidable loss. The gap is not a persona problem in the abstract - it is that Procurement's negotiation cycle and Finance's consumption data are never in the same room.
Note the pattern: the CoE has absorbed work that belongs to other personas. That is the classic centralised-team failure mode, and it scales badly.
2. The CoE wants to 'fix the showback'. What would you change so that engineering teams actually engage with it, and what would you refuse to add?
Read the worked answer
Change:
- Reconcile or label. A report that does not tie to the P&L is discarded on sight. Either reconcile to the finance actuals or show the gap explicitly as a line item. Trust is the prerequisite, not the nice-to-have.
- Allocate at the level of the decision. A team can act on 'your ETL job costs $4,100/month and 70% of that is re-scanning the same partition'. It cannot act on a department-level number. Move from department to service or workload, and attach the top three cost drivers per service.
- Make it a conversation with a name on it. Monthly 20-minute review per team, owned by the team, with the CoE in an advisory seat. The team brings the explanation; the CoE brings the data.
- Tie it to something they already care about - release cadence, latency, on-call load - so cost joins an existing trade-off rather than arriving as a separate mandate.
Refuse:
- More report formats. The problem is not the artefact, it is that no one has been asked to act on it.
- A new tool before the reconciliation and allocation logic is settled. You will spend a quarter on integration and inherit the same credibility gap.
- Chargeback. Do not move to hard chargeback while the data is contested; you will convert a reporting dispute into a budget dispute and lose the engineering audience entirely. Showback, then chargeback once the numbers survive two or three cycles of argument.
3. Who should own a Salesforce renewal that comes up in six weeks at $740K a year? Justify the split of responsibility.
Read the worked answer
No single persona. Assign it as a joint decision with distinct roles:
- Procurement owns the commercial negotiation and the contract - price, term, tier, true-up mechanics, exit and data-portability terms. That is their craft and it is the only seat with the negotiation leverage and the legal review process.
- FinOps Practitioner supplies the consumption and cost evidence - seats provisioned versus seats active, usage trend, per-user cost versus comparable tools, what the internal charge is today. Without this, Procurement negotiates on list price, which is exactly the current failure.
- Product Owner (the business owner of the tool) owns the requirement - which capabilities are actually used, what would be lost by dropping a tier, and whether the renewal should happen at all.
- Leadership owns the escalation if the business owner and Procurement disagree on scope.
The judgement call is that Procurement plus Finance alone will optimise the wrong thing (unit price on a contract that may be 30% over-provisioned), and Product alone will renew out of risk aversion. Sequence it: usage evidence in week 1, requirement validation in week 2, negotiation brief in week 3, negotiation weeks 4-6. Six weeks is enough if the data exists on day one - which is the argument for building the SaaS consumption view before the next renewal, not after it.