Pulse surveys during a merger: measuring the first hundred days.

The weeks after a merger is announced are the ones in which people decide what it means for them, and they decide on whatever information they have. Leaders, meanwhile, are working from the deal timetable, a communications plan and what they hear in the corridor. A merger pulse closes that gap. It is short, it is repeated, and because it asks the same questions each time it shows whether the mood is moving and where. This guide sets out how we would design and run one, what to read in the results and what to do between waves.

What a merger pulse is for, and what it is not

It is not an engagement survey. It does not try to measure the whole of working life; it measures how people feel about the joining together, whether the communication about it is landing, and where the anxiety concentrates. Keep that discipline. Every question added to a merger pulse that is not about the merger lengthens it, lowers the response and blurs the reading. The annual survey can wait; the pulse cannot.

It is also not a one-off. The value is in the trend. A first reading tells you what share of the words people chose were negative, which is interesting; the second tells you whether that is rising or falling, which is useful; the third tells you whether what you did in between worked.

Design the instrument once, then leave it alone

• How people feel: words, not a scale. Offer a fixed list of around twenty words, balanced across positive, neutral and negative, and ask people to choose up to three that describe how they feel about the merger. Illustrative words: hopeful, encouraged, confident, valued, enthusiastic; unsettled, worried, under pressure, flat, irritated; and a handful of neutral ones such as curious, undecided or waiting to see. Report the share of all words chosen that were positive, neutral and negative, and the net of positive minus negative. Then watch which words move. “Unsettled” falling while “flat” rises is a different situation from both falling, and a single satisfaction score would show neither.
• How it is being run: a handful of statements. Five agree/disagree statements, on a five-point scale, about the process rather than the outcome: whether people know where to take a question about the merger, whether the updates they receive are useful, whether they feel informed about what it means for them, whether they have confidence in how the integration is being run, and whether they feel supported. Resist the temptation to add a sixth answer option such as “I don’t have a strong opinion”. It absorbs exactly the lukewarm middle you need to see. If two statements are near-identical, merge them before the first wave, because you will not be able to afterwards.
• What would help: one open question. “What one thing would make the biggest difference to you right now?” This is where the specifics live. Code the comments into themes, and code them twice, independently, so the themes are not one reading’s impression.
• About you: only what you will report at. Business unit, location, length of service and grade are usually enough. Every extra demographic is a reason for someone to doubt the anonymity, so collect only the cuts you intend to publish.

Once the first wave has run, the instrument is fixed. Changing a word in the list or the wording of a statement breaks the comparison, and the comparison is the product. If a new question is unavoidable, add it at the end and label it as new.

Timing: waves at milestones, not by the calendar

Keep each window short and identical: about ten working days, launched midweek, with reminders on fixed days. A longer window in one wave than another changes who answers and makes the trend harder to trust.

Running it so that people answer

Say who is running it and that it is independent. Say exactly what anonymity means in this design: individual answers are not seen by the organisation, and no group smaller than a stated size is reported. If the survey uses a personal link, explain why (so that reminders go only to people who have not answered) and what it does not do (identify answers to anyone in the organisation); somebody will otherwise ask, and if the explanation is not already there, the doubt spreads. Anonymity and confidentiality in employee surveys covers the promise; repeat it in every invitation, not only the first.

Report the response rate and the number of respondents every wave. The base will move between waves, and a shift in who is answering is itself information: a falling response among one group can be the first sign that group has stopped believing anything will change.

Reading the results

What to do between waves

The pulse earns its place only if something visible happens before the next one. Answer the specific questions, even when the answer is “not decided yet, and here is when it will be”; a dated unknown is better received than silence. Settle the people questions early: terms, grading, titles and progression, because until those are settled every other message is read through them. Use the channels people already rate to carry answers rather than reassurance. Create contact between the two organisations below leadership level, so that people meet their counterparts before the structure asks them to work together. Then, at the start of the next wave, tell people what changed since the last one and that this is why you are asking again. Leading people through change describes the reactions you will be seeing at each stage; the change reaction spotter is a quick way for managers to place an individual.

Pitfalls

For the mechanics of launching any survey, the implementation plan applies; for turning a wave into actions, from survey results to action. If you want the pulse run independently and reported wave by wave, WinningFormula includes pulse surveys, and this is the method we use.