Lubinpla

How to calculate repeat complaint rate as a chemical supplier

Summary

Repeat complaint rate is the share of closed complaints that come back for the same customer, product and symptom within a fixed window. The number means something only when four choices are fixed before counting and kept the same across periods: the same-problem key, the window, the denominator and the rule for closing a complaint.

  • A helpdesk repeat contact rate divides solved tickets where the same customer came back about the same issue within 7 days by all solved tickets, and call centers often use 14 or 30 days (eesel, 2026).

  • UK electricity distribution reporting guidance defined a repeat complaint as dissatisfaction with the same or substantially the same matter as a previously resolved complaint, and excluded cases resolved more than 12 months earlier (Ofgem, 2012).

  • The same guidance treated a complaint as resolved when the customer made no further contact within 28 calendar days of the substantive response (Ofgem, 2012).

  • Ofgem's 2012-13 complaints metric gave the percentage of repeat complaints a 50% weighting in the overall score (Ofgem, 2013).

This definition fits business-to-business technical complaints about chemical products: performance, quality or documentation problems raised by industrial customers. It is an internal management metric, not a regulatory reporting definition. Where a regulator, certification body or customer contract defines repeat complaints, report to that definition and keep the internal rate separate.

What is the repeat complaint rate formula for a chemical supplier?

The formula takes the complaints closed in the period that recur for the same customer, product and symptom within the window, divides them by all complaints closed in the period, and multiplies by 100. Four choices sit inside that formula, and each one moves the result. Write them into the metric definition with an owner and a start date before the first report. Change any of them later only with a restated history.

The definition table below settles those choices before anyone counts. A rate built on undecided choices cannot be compared between months or between teams.

Choice

Option

What it does to the rate

Same-problem key

Customer only

Highest rate: any second complaint counts, even about another product

Same-problem key

Customer plus product

Middle: a new symptom on the same product still counts

Same-problem key

Customer, product and symptom

Lowest and most specific: counts only the problem that came back

Window

7, 30 or 90 days after closure (illustrative)

Longer windows catch slow failures and raise the rate

Denominator

Closed complaints, or complaining customers

A few customers who repeat often push the complaint-based rate up; repeats spread over many customers push the customer-based rate up

Closure rule

Attempted fix, or confirmed result

Counting attempted fixes as closures lowers the rate on paper

The key row matters most. A customer-only key turns an unrelated second complaint into a repeat, while the customer, product and symptom key counts only the problem that came back. For a chemical supplier the three-part key is the default; add the site when one customer runs several plants.

What counts as the same complaint when a chemical customer complains again?

A second complaint is the same complaint when it comes from the same customer about the same product and the same symptom, such as foaming, residue, corrosion or an out-of-specification result. A complaint about a different product, or a different symptom on the same product, is a new complaint, not a repeat.

UK reporting guidance defined a repeat complaint as one about "the same or substantially the same matter that was the subject of a previously resolved complaint" (Ofgem, 2012). In a 2010 industry working group, all energy suppliers agreed that a customer's different billing complaint a year after the first should not be counted as a repeat (Ofgem, 2010). A helpdesk guide makes the same point: match on customer and issue, not customer alone (eesel, 2026).

Chemical complaints drift on symptom wording. Fix a short symptom list per product family, agreed with the technical team, and code each complaint against it at intake. Record the lot or batch as well, so a repeat can be traced to the same lot or ruled out.

Which time window should a repeat complaint rate use?

The window should be at least as long as the product needs to show, in the customer's process, whether the fix worked. Published windows range from days to a year, because each was set for a different kind of problem.

Helpdesk practice uses 7 days as a default for email and chat, with call centers often using 14 or 30 days (eesel, 2026). One distribution company's published business rule counts two complaints within 30 days that are both chargeable to performance (PCF, 2018). UK electricity distribution guidance used 12 months after the resolution date (Ofgem, 2012), and one energy supplier reported agreeing a one-year timeframe with the Ombudsman (Ofgem, 2010).

For chemical products a 7-day window misses most real repeats, because a process chemical can take weeks to show that the problem is back. A 90-day window, illustrative, suits many process chemicals. Use a shorter one only for problems that show at once, such as a wrong label or a missing certificate of analysis.

Decide where the window starts. Ofgem counted from the resolution date (Ofgem, 2012), while the distribution company's rule uses the complaint date (PCF, 2018). Starting at closure measures whether the fix held.

What should the denominator of a repeat complaint rate be?

Use all complaints closed in the period as the denominator, because the metric asks how often a closed complaint comes back. A helpdesk repeat contact rate uses the same base, all solved tickets in the period (eesel, 2026).

A second base, complaining customers, answers a different question: how many customers had to complain twice. Report it beside the complaint-based rate when a few large customers generate most complaints, since one plant complaining five times can move the complaint rate on its own.

Three rules stay fixed along with the denominator. First, count each complaint once, in the period it closed, and count a repeat against the original complaint's period. Second, decide which complaints are excluded and write the exclusions down. The distribution company's rule counts only complaints chargeable to performance and leaves out certain coded complaints (PCF, 2018); a chemical supplier might exclude freight damage handled by the carrier, but the list must stay the same in every period. Third, report the raw counts with the rate, so readers can see how small a sample is.

Why do published repeat complaint rate definitions differ?

Each published definition was written for one kind of service, and problems take different times to come back in different services. Helpdesk tickets reappear within days; network repairs and product performance problems can take months. The comparison below shows where four published definitions part ways.

Source

Same-problem test

Window

Window starts at

Helpdesk guide

Same customer, same issue

7 days default; 14 or 30 in call centers (eesel, 2026)

Ticket solved

Distribution company rule

Two complaints chargeable to performance

30 days (PCF, 2018)

Complaint date

UK electricity distribution

Same or substantially the same matter

12 months (Ofgem, 2012)

Resolution date

UK energy suppliers, 2010 working group

Same person and topic is not enough: a different billing complaint is not a repeat

One year, agreed by one supplier with the Ombudsman (Ofgem, 2010)

Not stated

The definitions also differ on what "resolved" means. Ofgem's guidance counted a complaint as resolved when the customer was satisfied, agreed that all reasonable action had been taken, or made no further contact within 28 calendar days of the substantive response (Ofgem, 2012).

Before borrowing a benchmark, check that its key, window and closure rule match yours. If any of the three differs, the numbers cannot be compared. Take the same-issue test and the closure rule from these sources, then set a window that fits your products.

How do you use repeat complaint rate on an illustrative month of complaints?

To work the rate on a month of complaints, fix the key, the window and the denominator first, then count the closed complaints that came back.

In an illustrative March, a supplier closes 40 complaints from 28 customers. With a 90-day window from closure and the customer, product and symptom key, 6 of the 40 recur by the end of June. The repeat complaint rate is 6 divided by 40, or 15%. Those 6 repeats come from 5 customers, so the customer-based rate is 5 divided by 28, or 18%.

The key changes the illustrative answer. With a customer-only key, 10 of the 40 have a second complaint of any kind within 90 days, giving 25%. Four of those concern a different product or symptom, so in this illustrative month the customer-only rate overstates the repeat problem by 10 percentage points.

So does the closure rule. Suppose, as an illustrative case, 5 of the 40 were closed when a corrective action was proposed, before the customer confirmed the result. Two of those come back. Counting them as repeats is correct. Removing them because the case was "still open" hides the two failures the metric exists to show.

Report the March rate only after the 90-day window closes.

Why should attempted fixes be kept apart from confirmed results in repeat complaint counts?

A complaint closed on a proposed action can come back as a "new" complaint and never show as a repeat. Attempted fixes and confirmed results therefore belong in separate counts.

UK reporting guidance drew the same line. A complaint was not to be treated as resolved until the customer was satisfied, or reasonably believed to be satisfied, with the outcome, including awaiting the results of any monitoring before closing, and a course of action agreed but not yet completed was not to be recorded as resolved (Ofgem, 2012). For a chemical supplier, monitoring means the customer's next batch or run after the change.

Give each closure one of two codes, action proposed or result confirmed, and report the repeat rate for each code separately.

If complaints closed on a proposed action repeat far more often, the issue is closure discipline, not product quality. Delays inside your own team, such as waiting for a lab result or a reply from the manufacturer, are a different problem. Measure them as cycle times, as described in how to find where technical support cases get stuck, not as repeats.

Lubinpla's technical support performance analysis fixes the metric definition first, uses repeat customer incidents as evidence of how cases were solved, and reviews internal bottlenecks separately.

Frequently asked questions

What is a good repeat complaint rate for a chemical supplier?

No published benchmark fits chemical suppliers, because each organization sets its own key, window and closure rule. A helpdesk guide makes the same point for support teams (eesel, 2026). Track your own rate by quarter, by product family and by closure code, and look for a falling trend.

Does ISO 10002 define how to calculate repeat complaint rate?

ISO 10002:2018 gives guidelines for complaints handling, including analysing and evaluating complaints to improve the quality of products and services (ISO, 2018). Its public abstract sets no repeat rate formula, so the supplier fixes its own definition and keeps it stable.

Can a falling repeat complaint rate hide a problem?

Yes. Customers who stop complaining because they gave up, or who switched supplier, also lower the rate. A helpdesk guide recommends comparing the repeat rate with customer satisfaction measures so you know repeats fell because issues were fixed (eesel, 2026). ISO 10004:2018 gives guidelines for monitoring and measuring customer satisfaction (ISO, 2018).

Related posts