Reputation Trends Are Rarely Sudden

When a reputation score moves, the first instinct is to find the event that caused it. That's often the wrong place to start. Meaningful reputation change tends to build over several periods, and the event that draws attention may only mark when the movement became visible.

Start with the movement instead. Name the window you're reading it in, test how strong the evidence is, and only then check whether an event fits.

Movement often builds before it shows

Our advisory team sees this often in client readouts. Two things make change gradual, and a third makes it look sudden anyway:

• The first is exposure. A story can't shift the views of stakeholders who never encountered it, and a large share of any audience never sees a given story. Those who do see it read it through what they already believed, so the same story moves different people by different amounts.

• The second is depth. What people think about a company, across the specific factors they judge it on, sits underneath how they feel about it overall. An emerging issue can appear first in a single factor, one stakeholder group or one market, and sit there for several periods while the overall score holds. One pension fund we worked with posted a stable reputation that led its sector. Underneath, scores among existing members fell while scores among non-members rose. The total held because two opposite trends canceled out, and the falling group was the fund's own members.

• The third is spacing. Readings arrive at intervals, so gradual change tends to surface as a step. It stays out of view until the headline score finally moves or the next study arrives. With an annual study, a year of slow drift lands as one jump on the day the results come in.

Some movement really are sudden, and the event does explain them. They tend to carry three marks:

• They’re large.

• They land in a month when most of the audience met the same news.

• They show up on the factors the event relates to.

As an example, a major network outage at a telecom provider would fit all three: it reaches many customers at once, it's widely reported, and it lands on how people see the service. A movement without those marks is more likely to be gradual change surfacing.

How you measure your reputation helps you diagnose which is which, and what to do about it.

The window you pick writes the headline

The window decides whether a movement reads as a crisis or a blip, so it's the first thing to settle. Take the outage. In a monthly view, the drop that follows it can look like a crisis. In a quarterly or yearly view, the same drop may barely register. Both readings are accurate. Which one leadership sees depends on the window in the deck.

Measuring monthly lets you read the same movement through more than one window. RepTrak fields its survey every month and reports results by month, quarter, six months and year, all from the same readings. That means the outage month can sit beside the quarter and the year around it. If the longer windows hold steady, the drop was probably a blip. If they start bending the same way, it may be the start of something longer-lasting.

Two habits make the windows useful. First, choose the main window to match the decision the reading feeds. Monthly suits spotting what needs a closer look. Quarterly or six-month suits resetting priorities. Yearly suits planning and peer comparison. Second, read one longer window beside it, so a sharp month is always seen against the trend around it.

Direction, significance and persistence call for different responses

Before reacting to a movement, test it three ways: direction, significance, and persistence:

Each test supports a different default response. Direction is worth noting and watching. Significance is worth diagnosing. Persistence makes the strongest case for changing the plan, the budget or the message.

Compass, RepTrak's reporting platform, marks the first two on the changes it reports. A triangle shows the direction of a change. A filled triangle means the change is statistically significant, and an unfilled one means it isn't. The fill matters more than the size. A large move in a small segment can fail the significance test that a modest move across the whole audience passes, because the segment's sample is smaller.

The responses in the table are defaults. The outage carries all three marks of a real event, so it gets an event response straight away. A single significant change can also justify action when it sits on a heavily weighted driver or with a stakeholder group the business depends on.

Labeling each movement by the tests it passes also protects frequent measurement. Without a label, a monthly movement gets acted on, reverses the next month, and monthly data gets written off as noise. What failed was the presentation: a movement shown without its evidence.

Persistence can arrive by two routes. A significant change can hold, as the outage's drop might. Or a run of small moves in the same direction, none significant on its own, can add up across a longer window to a change that is. That second route is how gradual change surfaces, and a search that starts from events tends to miss it.

Catching it takes continuity from one report to the next. Every wave, our advisory team gives clients insights on what changed and why, with recommendations and actions to match. Each wave builds on the last and is tied to the goals the client set. So a persistent theme stays in view across reports.

Agree on the standard before the next results arrive

It's natural to start with the month's events and look for the one that fits the number. Events are memorable, and they're what communications teams manage. The trouble is that a search starting from events almost always finds one, whether or not it explains the movement.

Reverse the order. Start with the movement: when it began, which factors and audiences carry it, and which tests it passes. Then check whether an event carries the three marks. A label on every movement, naming its window and its tests, is what turns monitoring into prioritization. It decides what gets attention and what gets budget.

The practical step takes an afternoon. Add two lines to the top of your standing reputation readout: the window, with a longer one beside it as a check, and the tests each movement passes. Send the revised template to the executive team before the next quarterly results land, with one sentence on why. That way the standard is agreed on while nothing is moving, rather than set by the first inconvenient number.

RepTrak measures both layers, by stakeholder group and market. The Reputation Score captures how a group feels about a company. Seven drivers and 23 factors capture what it thinks.