First 90 Days as Head of Consumer Insights — July 2026

Jul 7, 2026 by Ethan Pidgeon

Taking on a new head of consumer insights role is one of those jobs where the first call you make matters more than the first deck you build. Before you touch a vendor contract or redesign a tracker, you need to know which stakeholders trust the data, which ones don't, and where the same question keeps getting asked because nobody believed the last answer. That's the map your first 90 days should produce.

TLDR:

• Spend your first 30 days listening and collecting artifacts before writing a single recommendation.
• Map stakeholders into three groups (Producers, Consumers, Affected) and sequence conversations by decision proximity.
• Data silos are mostly an ownership problem: roughly 68% of CPG marketing leaders name organizational silos as the top hurdle, per BCG 2025.
• Your day-60 deliverable should hit three formats (CMO deck, CFO Excel, brand brief) with the same evidence and confidence tiers in each.
• Merciv queries across social, cross-retailer reviews, syndicated research, and internal decks in one environment, with every claim carrying a source and confidence tier.

Days 1 to 30: Listen Before You Lead

The instinct on day one is to redesign something. Resist it. The first month is field research, and you are the researcher.

Book thirty-minute conversations with everyone whose work touches CPG consumer insights: the CMO, brand leads, category managers, the analytics director, procurement, IT, agency partners still on retainer. Ask three questions and shut up. What decision are you making next quarter that you wish you had better evidence for? Where did the last insights deliverable land, and what happened after? What do you not trust in the current setup?

While the calendar fills, collect artifacts:

• The last four QBR decks and every active tracker with its wave cadence
• Vendor contracts with renewal dates and the Slack channel where research requests get thrown over the wall
• Old brand health reports and any dashboard someone built and quietly stopped updating

You are hunting two patterns. Where does the same question keep getting asked because nobody trusts the last answer? And where does an answer already exist that nobody knows how to find?

Write nothing down as a recommendation yet. You have not earned the right to prescribe.

Mapping the Stakeholder Field

Stakeholder mapping is the second half of the listening month. Every person you booked lives in one of three groups, and you sort them fast.

Sequence conversations by decision proximity. The brand GM launching in Q3 outranks the category manager mid-cycle. Listen for two things: the language they use to describe consumer questions, and the moment they trail off because the answer never came. That gap is your map.

You are not building an org chart. You are building a flow diagram of trust.

Auditing the Existing Insights Stack

By week three, the artifacts you collected have a shape. Lay them out as a stack audit before you touch a vendor call.

Four layers to inventory:

• Syndicated subscriptions: what is active, what it covers, what it costs, when it renews
• Listening and review tools: Brandwatch, Sprinklr, Meltwater, Talkwalker seats, plus cross-retailer review pulls happening in spreadsheets
• Internal repositories: the shared drive, SharePoint, Looker dashboards nobody opens, Snowflake tables the analytics team owns
• Connective tissue: what talks to what, and where the joins still happen by hand in Excel on Friday afternoons

A rule of thumb circulated in CPG circles places the target at roughly one insights professional per 200 to 400 million in revenue via this staffing reference, depending on category complexity. If your inherited team looks under-resourced against that ratio, the audit is about finding what the current stack cannot answer, not buying more consumer insights platforms.

Write two columns: what you own, and what it actually delivers. The gap is your Q2 agenda.

Days 31 to 60: Define the Insights Charter

Month two turns listening into a document. Not a strategy deck. A one-page charter naming what insights is on the hook for, and what it will politely refuse.

Draft the scope first. Three or four market research techniques and methods the team owns end to end: brand health, category and competitive read, launch validation, SKU-level consumer perception. Everything else routes through a written triage rule back to the requester with a suggested next step.

Then write the operating principles short enough that a brand GM will read them:

• Every finding carries a source, date, and confidence tier (high, directional, exploratory).
• Standard reads turn in five business days; deep investigations get scoped in writing first.
• Anything above the CMO gets a second pair of eyes before it ships.

Socialize in two passes. Walk it one on one to the producers and consumers you mapped in month one and let them redline. Then bring the revised version to your boss with tradeoffs named out loud: what stops, what slows, what sharpens. Call it version 0.9 through month two. A charter edited in public is trusted more than one that arrives finished.

Diagnosing Data Silos and Synthesis Gaps

The synthesis gap has a signature. Social lives in a social listening tool that ignores internal data one team touches; syndicated pulls sit in a shared drive analytics owns; research decks age on SharePoint; POS lands in Looker with no link back to sentiment. Industry estimates suggest teams lose roughly a third of weekly work hours reassembling feeds by hand.

The instinct is to call it a tooling problem. It usually is not. Roughly 68% of CPG marketing leaders, per BCG's 2025 report, name organizational silos as the biggest hurdle, meaning joins fail because ownership is scattered before data moves.

Diagnose accordingly:

• Where does the same question route to three owners with three answers?
• Which feed has no named steward, so nobody updates it and everyone still cites it?
• What synthesis step (the work of connecting internal data to external consumer signal) happens in one analyst's head every Friday, and what breaks the week they take vacation?

Symptoms are duplicate research and slow readouts. Root causes are unassigned ownership and a synthesis layer living in tribal memory.

Building Cross-Functional Relationships That Stick

Relationships that stick are built on a calendar, not at offsites.

Set up a recurring touchpoint for each partner, matched to what they actually need:

• Finance: a monthly 30-minute methodology walkthrough for any number heading to the CFO. Show the confidence tier and the source.
• Brand: standing seat in the weekly brand meeting for SKUs closest to launch. Speed beats depth here.
• Sales: a shared doc updated before every top-ten buyer meeting with two competitive reads and one shopper signal.
• Category: a quarterly shelf-level review tied to the retailer's category review cadence, not yours.

Once finance cites your confidence tiers in a board deck without prompting, you have achieved leadership buy-in for consumer insights strategy.

Days 61 to 90: Deliver Proof and Build a Mandate

By day sixty, you have earned the right to ship. What you ship matters more than how fast.

Pick a mandate-builder over a quick win, even if it takes two extra weeks (a cross-functional SKU discontinuation analysis, for example, beats a one-off brand sentiment pull as your day-60 showcase). Triangulate three inputs:

• A next-quarter decision with real budget (line review, launch gate, portfolio pruning)
• A question two stakeholders in different functions raised in month-one interviews
• A synthesis your current stack cannot produce without a week of manual joining

Then produce one artifact in three formats, same evidence, same confidence tiers:

Ship it with a footer naming when the next update lands and what will change. That footer is the mandate. It tells the organization insights runs on a cadence now, and every subsequent deliverable inherits the format expectations you just set.

Making Insights Defensible to Leadership

Defensibility is what determines whether a finding leaves your inbox and enters a decision.

Four attributes make an insight survive a CFO's follow-up:

• Source named inline, not buried in an appendix
• Confidence tier stated before the recommendation
• Retrieval date on every number, so leadership knows what "current" means
• Sample validity called out when the n is thin, in the same font as the finding

Format for the room. On an executive slide, the recommendation sits on line one; the evidence stack below; the confidence tier in the header.

Measure influence with citation rate: the share of QBR decks, brand plans, and capital requests that reference your team by name. In our work with CPG insights teams, once that number crosses roughly 60 percent, the function has moved from filing reports to shaping decisions.

Building a Research Governance Framework

Governance is what turns a person into a function. Without it, every decision reruns through your judgment.

Four documents make the framework operable:

• Methodology playbook: standard designs for the four question types the charter owns, with minimum sample sizes and analysis steps named
• Confidence-tier rubric: High needs three recent sources in agreement; Directional accepts thin but aligned signal; Exploratory flags a single-feed read
• Source quality standard: named provider, retrieval date within 90 days for market claims, verbatims quoted with n and date
• Triage rule for returns: requests with no decision date, no budget, or a sample the design cannot support get sent back with a scoped counterproposal, not a polite yes

Publish it. Revise quarterly.

How Merciv Helps Insights Leaders Build the Foundation Faster

The synthesis gap you diagnosed in week three is what we built Merciv for. Query across social, cross-retailer reviews, licensed syndicated research, and internal decks in one environment from day one, with every claim carrying a source, retrieval date, and confidence tier.

Outputs land in the formats your charter already named:

• A ten-slide CMO deck with recommendation on slide one
• An Excel with a confidence column for finance
• A one-page brand brief with so-what and now-what

No SQL, no Python, no Friday spent joining exports by hand. Research cycles that ran weeks compress to days. SOC 2 Type II, zero training on your data by policy, and tenant isolation at the infrastructure level keep procurement from stalling the pilot.

Final Thoughts on Setting Up a Consumer Insights Function That Lasts

Ninety days goes fast, and the temptation to show quick results is real. But the leaders who build lasting influence do it by mapping trust before they ship answers. Your charter and your confidence tiers become the vocabulary your organization borrows. Take a look at Merciv for enterprise if you want to see how the synthesis layer runs in practice.

FAQ

What should a new Head of Consumer Insights do in the first 90 days before touching any tools or vendors?

Listen first, diagnose second, act third. Spend days 1 through 30 running structured conversations with every stakeholder who touches insights work, collecting QBR decks, vendor contracts, and any tracker with a wave cadence, and mapping where the same question keeps getting asked because nobody trusts the last answer. The audit you run in week three shapes everything that follows, including which tools deserve renewal and which synthesis gaps are actually ownership problems, not tooling problems.

How do I make consumer insights defensible to a CFO or CMO who will pressure-test my findings?

Every finding that reaches a CFO needs four things stated inline, not buried in an appendix: the source named in the sentence, the confidence tier declared before the recommendation, a retrieval date on every number, and a clear callout of sample size when the n is thin. Format for the room: put the recommendation on line one of the executive slide, the evidence stack below it, and the confidence tier in the header so leadership can see your work without asking for a follow-up.

What's the fastest way to build cross-functional trust as a new insights leader?

Skip the offsite and put it on the calendar instead. Match the touchpoint to what each partner actually needs: a monthly 30-minute methodology walk-through for Finance before any number heads to the CFO, a standing seat in the weekly brand meeting for SKUs closest to launch, and a shared doc with two competitive reads and one shopper signal updated before every top-ten buyer meeting with Sales. The signal that you are embedded is when Finance cites your confidence tiers in a board deck without prompting.

Can I use ChatGPT or Claude to produce cited consumer insights for a leadership readout?

General AI tools are genuinely useful for drafting questionnaires and coding open-ends, but they hit a hard limit when the output needs to survive a CFO's follow-up. They have no access to licensed syndicated research, no source attribution, no confidence scoring, and no audit trail, and run-to-run outputs can shift materially on the same prompt. Merciv is built for the synthesis step that follows: social, cross-retailer reviews, licensed syndicated data, and internal decks queried together, with every claim carrying a source, retrieval date, and confidence tier your CMO can click through and verify.

Meltwater vs Brandwatch vs Merciv for a new Head of Consumer Insights building out a defensible stack?

Meltwater and Brandwatch are social-listening-first tools that do their designed job well: surfacing consumer conversation at scale. The ceiling appears when the question moves from "what are people saying" to "where is my category moving and why did my SKU lose shelf space," because neither joins social signal to syndicated data, cross-retailer reviews, or internal research in a single query. Merciv is positioned differently: it treats social as one input among several and delivers outputs with source attribution and confidence scoring across all feeds, which matters most when findings need to hold up in front of a board or a buyer.