Marketing mix modelling (MMM) & Testing Services
Assemble and validate test data — geo-level spend and conversion series, checking pre-period parity between treatment and control, spotting the coverage gaps that invalidate a design before it launchesSupport test design under review — market matching and control selection, power and minimum detectable effect sanity checks, and identifying contamination risks such as geo-targeting settings that don't behave the way the platform's documentation claimsRun analysis and read the results honestly — pre-treatment fit diagnostics, lift estimates with their intervals, and what a null result does and doesn't tell youQualify client data for MMM — spend coverage across channels, whether there's enough variation in spend to identify an effect at all, series length and granularity, collinearity between channels, and gaps that will bias the resultAssemble and validate model input datasets, and investigate the discrepancies that surface when you doSupport model runs and read the diagnostics — fit, residuals, convergence, and whether a channel's estimated contribution is plausibleContribute to output-extension work under review — building on an existing MMM result, for example forecasting or budget scenario work derived from itCompare results across methods — where MMM, incrementality, and platform-reported figures disagree, understanding why is the interesting part of the jobModel trust and diagnostics
First and second line on client trust queries — investigating why a number changed, working in SQL against client data to isolate the causeDistinguish a bug from a methodology change — attribution window changes, model recalibration, data feed gaps, and platform reporting shifts all look similar from the outside and have very different signatures underneathTriage PSPs on model trust, resolve what you can, and escalate what turns out to be a genuine model problem with a clear diagnosis attachedReconcile platform-reported figures against our measurement — why walled-garden ROAS disagrees with ours is the hardest recurring question in the business, and you'll be learning it from the insideLog and tag incidents consistently, so recurring failure patterns become visible and can be automated away rather than repeatedly handledClient communication and enablement
Run templated explainer sessions under review, walking clients through how our measurement worksDraft documentation and presentations above the core explainer content, and feed recurring query themes back into the source materialFact-check methodology claims in product marketing collateral before it goes outLevel doesn't gate what you're allowed to attempt. It scales how much support you get doing it. A Graduate can work on MMM output-extension work; it just carries heavier review than it would for a Mid.
Incrementality delivery: Triages test type and routes to Product where self-serve applies. Prepares and validates test data. Supports design and analysis under review.
MMM delivery: Qualifies client data. Assembles model inputs. Contributes to output-extension work under heavy review.
Model trust & PSPs: First and second line on trust queries. Triages PSPs, attempts resolution, escalates genuine model problems with a diagnosis.
Client sessions: Runs templated explainer sessions under review.
Documentation & enablement: Drafts documentation and presentations above the core explainer. Feeds recurring query themes back into explainer content. Fact-checks methodology claims in product marketing collateral.
Automated trust workflows: Logs and tags trust incidents consistently so recurring failure patterns are visible. Surfaces themes from the query queue into workflow design.
Supervision: Detailed instruction; works closely with senior colleagues. Development time protected against product and consultancy workload.
Upskilling is funded, not assumed. A named share of entry-tier capacity is held for development work and protected against the support queue. Nobody upskills in the gaps between tickets, so we don't pretend otherwise.
The step to Junior is independence: running templated sessions and first-line triage unaccompanied, and resolving the full range of inbound queries with only moderate senior input.
The step to Mid runs through one of two routes — owning a piece of MMM output-extension work, or taking a recurring trust problem you already handle manually and turning it into automated monitoring that alerts an account manager before the client notices. Both are genuinely technical, and both are allocated deliberately rather than first-come-first-served. From Mid, you own MMM and incrementality delivery with senior sign-off.
Why Fospha
Causal measurement is the product. Geo experiments, MMM, difference-in-differences, and Bayesian attribution are what the business sells. You'll be working on them, not adjacent to them.Live client work early. You'll be on real test and MMM delivery in your first months, with review rather than distance.Structured support, not sink-or-swim. Session allocation rules, a maintained escalation list, and protected development time exist so that being new isn't a liability.A published ladder. You'll know what the next level requires and what evidences readiness for it from your first week.By submitting your CV you understand that we have a legitimate interest to use your personal data for the purposes of assessing your eligibility for this role. This means that we may use your personal data to contact you to discuss your CV or arrange an interview, or transfer your CV to the hiring manager(s) of the role you have applied for. You can ask us at any time to remove your CV from our database by emailing recruitment@blenheimchalcot.com – but please note that this means we will no longer consider you for the role you have applied for. You can review our privacy policy here.
