
40% of customers stop buying after poor service
Poor customer service does not just cost a complaint to handle. It often costs the customer. In 2020, 40% of surveyed US consumers said they had stopped buying from a company because of disappointing customer service.
This figure, from Statista's dossier on customer service in the US, makes more sense when you put it next to three other data points from the same dossier: how much customers value service, how fast expectations are rising, and how often customers get in touch.
Four numbers that echo each other
First: 40% of consumers stopped buying from a company after poor service (2020). This is not an intention; it is reported behavior.
Second: 66% considered customer service very important when choosing or staying loyal to a brand (2018). Third: 56% said their expectations were higher than a year earlier (2018).
Fourth: 58% had contacted customer service in the past month (2020). Customer service is therefore not a niche topic involving a few unhappy customers. It concerns more than one customer in two, every month.
Expectations rising faster than resources
Together, these numbers describe a tension. Customers contact service often, expect more and more, and are ready to leave if they are not satisfied. Meanwhile, customer service teams work with headcount and tools that do not evolve at the same pace.
In many companies, quality is still measured with a handful of indicators: response time, first-contact resolution, a satisfaction score after the conversation. These indicators are useful, but they say little about what the customer felt, or about what pushed them to leave.
There is also a comparison effect. Customers judge a service against the best experiences they have had elsewhere, not just in the same industry. A fast, precise answer from an online store becomes the new standard they expect from their bank or phone provider.
This is why the 56% figure matters: rising expectations do not wait for budgets to follow. Teams that keep doing the same thing slowly fall behind what customers consider normal.
Why scores are not enough
A satisfaction score sums up an experience in one number. It does not say whether the customer was annoyed by the wait, by the answer, or by having to repeat their problem. Nor does it say whether that customer has already had three similar contacts this quarter.
Yet it is often the accumulation that triggers the departure. One disappointing conversation can be forgiven. Three disappointing conversations on the same topic, much less so. That accumulation is visible in customers' words long before it shows up in cancellation statistics.
We looked at this gap between scores and review content in our analysis of 200 G2 reviews.
Listening across every contact
If 58% of customers contact service every month, the volume of conversations is considerable, even for a mid-sized company. Automatic language analysis can read each of these conversations and extract three things: the topic, the emotion expressed and the risk of leaving.
Teams can then focus their energy on the customers and topics that matter most, instead of treating every contact the same way. Managers get an overview that neither scores nor occasional call listening can provide.
In practice, a manager can see every Monday the three topics that generated the most negative messages the previous week, the list of customers who chased several times, and how the overall tone is evolving. That review takes a few minutes, compared with hours of manual reading for a partial result.
The analysis relies on models such as Mistral, BERT, Llama, GPT or Claude depending on the case. Data stays hosted in France.
The takeaway
The data is American and dates from 2018 and 2020, but the mechanism is universal: disappointing customer service drives customers away, and those customers say so before they go. Companies that systematically read what their customers say are a step ahead of those that settle for an average score.
A practical first step is to take the conversations of the last month and answer one question: which three topics made customers most unhappy? If nobody in the company can answer quickly, that is the gap to close first.
Sources
- Statista (2025). Customer service in the U.S. — dossier.