How to Assess Post-Acquisition Customer Sentiment

How to Assess Post-Acquisition Customer Sentiment

Most post-deal sentiment work is too late, too thin, or too flattering. By the time churn shows up in the numbers, the customers who mattered most have already made their decisions. The question is not whether customers are unhappy, it is whether the segments that drive economic value are quietly preparing to leave, and whether you can detect that in time to act.

This guide sets out how to build an assessment that gives you signal early, separates noise from real risk, and translates sentiment into decisions a board can act on.

Start with the segments that matter, not the customer base

Aggregate sentiment scores are misleading after an acquisition. A stable overall NPS can hide a collapse among the top 5% of customers by revenue or lifetime value. Before you measure anything, define the three or four segments that account for the majority of economic value and relationship risk: high-value retail, advised clients, intermediary-introduced books, corporate relationships with cross-sell density. Run the assessment by segment from day one. If you only look at the average, you will miss the exit.

Set a pre-close baseline, even if it is imperfect

The single most common failure is having nothing to compare post-close sentiment against. Wherever possible, capture a baseline in the weeks before completion: existing NPS, complaint volumes and themes, call centre sentiment, social listening, intermediary feedback, and broker scorecards. If the target will not share customer data pre-close, build a proxy from public reviews, Trustpilot trends, FOS data, and industry benchmarks. An imperfect baseline beats none.

Use at least three independent signals

Surveys alone will not tell you the truth. Combine:

• Direct signals: targeted surveys, relationship manager debriefs, structured interviews with top-decile clients.
• Behavioural signals: login frequency, balance movements, product utilisation, direct debit cancellations, partial transfers out, intermediary redirection of new business.
• Unsolicited signals: complaints (volume, theme, escalation rate), social and review platforms, call recordings analysed for emotional tone, broker and adviser commentary.

Where the three diverge, the behavioural data is almost always the leading indicator. Customers say they are fine, then quietly move money. Trust the behaviour.

Run the assessment on a deliberate cadence

Week 2 to 4: complaint themes, call centre sentiment, intermediary pulse. You are looking for early operational damage, billing errors, login problems, branding confusion.

Month 2 to 3: structured interviews with priority clients and intermediaries. This is when relationship managers can still recover situations.

Month 4 to 6: full segmented survey against baseline, paired with behavioural cohort analysis.

Month 9 to 12: repeat, with focus on whether early issues have been resolved or have hardened into structural distrust.

What most acquirers get wrong

Three recurring errors. First, they outsource the assessment to the same agency running brand tracking, who produce reassuring averages. Second, they let the integration team mark its own homework, asking customers about things integration has not yet touched. Third, they treat intermediaries as a channel rather than a customer, when in advised and intermediated books the broker view often predicts retention better than the end-client view.

Good looks like this: a small, independent workstream reporting to the deal sponsor or board, with direct access to behavioural data, running against a defined baseline, segmented by economic value, and triangulating at least three signal types.

Translate sentiment into decisions, not dashboards

The output should not be a score. It should be a short list of decisions: which segments need a retention intervention, which integration milestones need to be slowed or resequenced, which intermediary relationships need partner-level contact, and where the deal thesis itself is at risk. If your assessment cannot drive a decision to pause a migration, redirect spend, or escalate to the board, it is not earning its place.

Your next step

Before the next integration steering committee, ask one question: what is our baseline, and which three signals are we triangulating by segment? If the answer is a single NPS number and a complaints log, you do not yet have an assessment. You have a dashboard. Fix that this quarter, while there is still time to act on what you find.