1. What the first-time fix rate measures - and what it does not
The first-time fix rate (FTFR) is the share of on-site visits after which the service case was closed: no second visit, no callback, no escalation. The formula is simple: visits resolved at the first visit, divided by all on-site visits, times 100. If 500 visits are made in a month and 380 of them are completed at the first visit, the first-time fix rate is 76 percent. The German customer service association KVD lists the metric in its programme for the Service Congress 2026 alongside response time, cost per visit and productivity as one of the central steering figures in technical service.
As simple as the formula is, it is just as easy to distort. Three decisions determine whether the number is worth anything:
- The window. When does a case count as resolved? The Aquant Field Service Benchmark Report 2024 (vendor study) measures the first-time fix rate at 30 days and warns explicitly: anyone measuring in seven- or 14-day windows overstates their own rate and understates resolution costs, because the second visit for the same problem falls outside the window and counts as a new case.
- Grouping follow-up tickets. Two “successful first visits” on the same machine within three weeks are one failed visit from the customer’s point of view. As long as tickets for the same root problem are not grouped, the rate is a fiction.
- How remote resolutions are treated. A case the hotline resolves on the phone is not an on-site visit and does not belong in the denominator. Whoever resolves the easy cases remotely improves the service - and may at the same time measure a falling first-time fix rate, because only the hard cases are left in the field. That is why FTFR and the remote resolution rate always belong on the same sheet.
One confusion is particularly common: first call resolution is a hotline metric and measures how many enquiries are settled at the first contact without a dispatch. The first-time fix rate measures on-site visits. The two have different denominators, different benchmarks and different levers. Whoever forces both into one number can no longer steer either.
2. The 2026 benchmark: where your service stands
The most robust publicly accessible dataset comes from a vendor: the Aquant 2026 Field Service KPI Benchmark Report (published 19 February 2026) analyses anonymised data from 161 service organisations: nearly 30 million service events, 7 million assets and 8.3 billion US dollars in service costs over three years. Aquant sells AI software for service and has an interest in wide gaps between good and poor organisations - the figures are therefore reference values, not neutral statistics. But there are no more neutral ones: neither the VDMA nor the KVD publishes a first-time fix benchmark.
| Metric (2026) | Bottom 20 % | Industry benchmark | Top 20 % |
|---|---|---|---|
| First-time fix rate | 60 % | 77 % | 88 % |
| Time to resolution | 10 days | 4.5 days | 2.5 days |
| Share of failed visits in total service cost | 44 % | 25 % | 14 % |
| First-time fix gap between top technicians and the rest | 10 points | - | 2.9 points |
| Workforce retention | 66 % | - | 87 % |
Source: Aquant 2026 Field Service KPI Benchmark Report (vendor study, 161 service organisations, press release of 19 February 2026). According to the same study, one in five cases could be resolved remotely and 20 percent of truck rolls are unnecessary.
Two figures in the table deserve more attention than the gap between 60 and 88 percent. First, the cost share: failed visits account for 25 percent of total service cost at the median and 44 percent at the weakest organisations. Second, the gap inside the company’s own team: at the best organisations the top technicians are only 2.9 points above the rest, at the weakest ten points. Aquant attributes this to knowledge that stays concentrated among a few experienced technicians - the same diagnosis the German studies in the next section arrive at.
For machinery manufacturers, the 2024 edition is more revealing because it breaks the data down by industry: the median for industrial machinery was 71.9 percent, against 76 percent across all industries. A company in mechanical and plant engineering at 72 percent is therefore at the industry average, not adrift - and still has a long way to go to the 88 percent of the best organisations.
3. Why repeat visits happen: parts, experience, information
Anyone who wants to improve the first-time fix rate has to know why the first visit fails. The only openly accessible survey that separates the causes is old but cleanly documented: in January 2013 the Aberdeen Group surveyed 156 service organisations for the analysis “Fixing First-Time Fix” (March 2013, analyst firm). As reasons for a second or third visit, 51 percent named missing or wrong spare parts, 25 percent a technician without the necessary experience and 13 percent insufficient time on site. The average first-time fix rate at the time was 75 percent; the best organisations reached 89, the weakest 56 - almost the same spread as 13 years later at Aquant.
At first glance this argues against the thesis that knowledge is the bottleneck: the most common reason is the part, not the information. But the same respondents named better diagnosis or triage at the dispatch or initial call level as the most effective improvement, at 58 percent - 25 points ahead of better field-based access to parts (33 percent), more intelligent scheduling (32 percent) and more training (25 percent). The wrong part in the van is in most cases the result of an incomplete diagnosis on the phone. The parts question is an information question, one step earlier.
What this feels like in the everyday work of German-speaking service technicians is shown by the Insight-Report Service 2022, which the documentation service provider kothes conducted with the German customer service association KVD and its partner associations in Austria and Switzerland (survey April to July 2022, 43.6 percent of respondents from mechanical and plant engineering; the report does not state the number of participants). The study asked directly about the first-time fix rate:
| How often did missing information mean that further visits were needed? | Share of respondents |
|---|---|
| Never | 5.3 % |
| In up to 10 percent of visits | 43.8 % |
| In 10 to 20 percent of visits | 24.8 % |
| In 20 to 30 percent of visits | 15.9 % |
| In 30 to 50 percent of visits | 7.1 % |
| In more than half of visits | 3.1 % |
Source: kothes/KVD/KVA/SKDV, Insight-Report Service 2022, survey of service technicians in Germany, Austria and Switzerland. For nearly 95 percent of respondents, missing information had at some point been the reason for a second trip; for a good quarter it happens on more than every fifth visit.
The 95 percent is the number that ends up on vendor slides. The distribution is the more honest statement: for the majority the information gap is an occasional problem, for a good quarter of technicians a constant one. That quarter decides the organisation’s rate - and it is usually not the quarter with the least experience but the one with the oldest machines and the rarest fault patterns.
The rest of the study explains where the gap comes from. More than 90 percent of respondents spend at least half an hour a day searching for information, more than a third at least an hour, and in mechanical and plant engineering 12.4 percent more than two hours. For nearly 60 percent it is unclear where to look at all. A good 83 percent get by with their own notes, half say step-by-step troubleshooting instructions are missing, and almost 44 percent receive information exclusively on paper. And the number every software concept has to answer before it starts: almost 70 percent have no internet access at the site.
4. What a repeat visit costs: a model calculation
There is no robust, dated source for the full cost of a service visit in euros - the figures circulating online come from vendor blogs without a methodology. What does exist are documented multipliers: according to Aberdeen (2013), a first visit that did not resolve the issue led on average to 1.6 additional dispatches; according to Aquant (2024), it takes 2.7 visits in total to resolve, adds 13 days and produces resolution costs 44 percent above the cost of a single work order. From that, a model calculation can be built into which you insert your own values:
| Item | FTFR 72 % | FTFR 77 % |
|---|---|---|
| On-site visits per year (assumption) | 6,000 | 6,000 |
| Not resolved at the first visit | 1,680 | 1,380 |
| Additional dispatches (1.6 per unresolved first visit) | 2,688 | 2,208 |
| Cost of additional dispatches at EUR 450 per trip (assumption) | EUR 1,209,600 | EUR 993,600 |
| Difference at five percentage points | 480 additional dispatches, EUR 216,000 per year | |
Model calculation, not a projection. Assumptions deliberately conservative: 6,000 visits correspond to an organisation with roughly 40 to 50 technicians; 72 percent is close to the Aquant median for industrial machinery (2024); 1.6 additional dispatches is the lower of the two documented multipliers; EUR 450 per trip (travel, labour, expenses) is a placeholder, not a study value - replace it with your full cost.
The table only counts what is visible in the service cost centre. Three items are not in it and are often larger in practice: at four hours per trip, the 480 dispatches tie up around 1,900 technician hours that are missing for maintenance contracts or new installations; the spare part that is ordered after all on the second attempt costs express shipping; and at the customer, resolution takes 13 days longer, which they remember at the next maintenance contract. Aberdeen found a clear relationship in 2013: at organisations with a first-time fix rate above 80 percent, customer retention was 88 percent; at organisations below 50 percent, it was 60 percent.
5. Six levers for a higher first-time fix rate
The levers follow the path of a service case: from intake on the phone through preparation and the visit to the service report. The first three act before anyone drives off - according to Aberdeen, that is where most of the problem originates. Where AI demonstrably helps and where it does not is described in more detail in the guide to AI in after-sales; here it is about the levers themselves, independent of the tool.
58 percent of the organisations surveyed by Aberdeen saw the biggest lever here, and the logic has not changed: every case logged with fault code, serial number, symptom description and the machine’s history sets off with the right part and the right technician. For that, the hotline needs access to the same sources as the field: documentation, service reports, the bill of materials of the specific machine. A knowledge system that searches these sources in the context of the serial number turns a dispatcher into a triage desk - provided it shows the source for every answer.
According to Aquant 2026, one in five cases could be resolved remotely, and according to the 2025 edition a third of all service interactions are simple informational queries. Every one of these cases that is not driven to disappears from the denominator of the first-time fix rate and from the cost centre. The lever is not software but a rule: no dispatch without a documented remote attempt, except where safety or contract require it.
51 percent of the organisations surveyed by Aberdeen named missing or wrong parts as a reason for repeat visits. The error rarely happens in the warehouse but at the question of which part is needed at all. Whoever evaluates the case history of the same machine and the same series sees which parts were actually replaced for which fault pattern in the past - and packs them before the technician discovers on site that the customer’s variant has a different valve. The prerequisite is a bill of materials per serial number that knows about modifications.
The technician in front of the machine needs the fault code table, the wiring diagram of the specific variant and the service report of the colleague who fixed the same fault on the same machine type three years ago. A good 83 percent get by with their own notes today, because the system in which this information should live either does not exist or cannot be reached on the shop floor. Almost 70 percent have no internet access at the site - a knowledge system that only works online does not help the technician in the customer’s basement. What such a system must be able to do for technicians to trust it is set out in the buyer’s checklist for AI assistants in service.
At the weakest organisations the top technicians are ten points above the rest, at the best only 2.9. The difference is not talent but whether the knowledge of the three most experienced people can be retrieved by everyone else. A quarter of the organisations surveyed named a lack of experience as a reason for repeat visits - and experience cannot be trained when the fault only occurs every two years. It can, however, be documented and found again if service cases are captured in a structured way.
Every resolved case is the answer to the next identical case - if it is captured so that it can be found again: fault pattern, cause, replaced part, duration, peculiarity of the variant. A report consisting of “fault fixed” is worthless for the first-time fix rate. A report with cause and cause code is the data basis for levers one to five. It is the least spectacular lever on this list and the only one without which the others do not work for long.
6. When a higher first-time fix rate is not the goal
The first-time fix rate is not a metric to be maximised. Aberdeen noted as early as 2013 that 100 percent is not the ideal state if the way there leads through larger van stocks or blanket parts replacement - somewhere there is a point at which the next percentage point costs more than it saves. In our view there are four situations in which the rate is the wrong first lever:
- The problem is parts logistics, not knowledge. If the diagnosis is right and the part is still missing because the warehouse does not stock it or the lead time is six weeks, no knowledge system helps. Then stocking by series and installed base is the topic.
- The rate is already above 85 percent. Then the remote resolution rate is usually the bigger lever: every case that is not driven to saves more than every case resolved at the first attempt.
- Cases rarely repeat. Whoever services a handful of special-purpose machines a year whose faults never resemble each other has no case history to draw on. There, experience and time on site count, not triage.
- Nobody measures. A first-time fix rate estimated from gut feeling cannot be improved, because nobody notices whether it moves. Aberdeen found in 2013 that 17 percent of organisations did not know their rate. In that case, section 7 is the only sensible first step.
And a warning against the reverse conclusion: the measures in section 5 are prerequisites, not guarantees. What reaches your own operation depends on data quality, installed base and usage. That is why your own baseline belongs before every decision about tools.
7. First steps: measure the first-time fix rate before you improve it
The measurement plan fits on one page and needs no software in the first weeks that you do not already have:
- Fix the definition and write it down. Denominator: all on-site visits. Numerator: visits after which no further visit for the same case on the same machine was needed within 30 days. Remotely resolved cases do not count in the denominator but in a rate of their own.
- Group follow-up tickets. Every ticket gets the machine’s serial number and, if it is a follow-up visit, the number of the first visit. Without that link the rate cannot be calculated.
- A cause code for every repeat visit. Four codes are enough: part (missing or wrong), diagnosis (fault different from assumed), information (documentation or history missing on site), time or access (machine not available, job too large). The technician sets the code in the report, not the dispatcher afterwards.
- Four to eight weeks of baseline. Overall rate, rate per series, rate per technician, distribution of cause codes. Only the breakdown shows whether you have a parts, a knowledge or a planning problem - and with it which lever from section 5 comes first.
- Three metrics together on one sheet. First-time fix rate, remote resolution rate and cost per resolved case. Each on its own can be dressed up; all three together cannot.
Whoever has run through this plan once knows after two months where the repeat visits come from. That is more than most vendor presentations know about your operation - and it is the prerequisite for a pilot with a knowledge system to deliver a decision at all.
8. Frequently asked questions
What is a good first-time fix rate?
According to the Aquant 2026 Field Service KPI Benchmark Report (vendor study, 161 service organisations), the industry benchmark is 77 percent; the top 20 percent reach 88 percent, the bottom 20 percent 60 percent. For industrial machinery, the 2024 edition reported a median of 71.9 percent. These values are only comparable with a measurement that uses the same window (30 days) and the same definition.
How do you calculate the first-time fix rate?
The number of on-site visits after which no further visit for the same case was needed within a fixed window, divided by the total number of on-site visits, times 100. Two decisions matter: the window (Aquant measures at 30 days; shorter windows overstate the rate) and the rule that follow-up tickets for the same case are grouped together.
Is the first-time fix rate the same as first call resolution?
No. First call resolution measures how many enquiries the hotline resolves at the first contact without anyone being dispatched. The first-time fix rate measures how many on-site visits are completed at the first visit. The two metrics have different denominators and cannot be compared with each other.
What is the most common reason for a repeat visit?
In the only openly accessible survey that separates causes (Aberdeen Group, 2013, 156 service organisations), 51 percent named missing or wrong spare parts, 25 percent a technician without the necessary experience and 13 percent insufficient time. The same respondents named better diagnosis or triage at the initial call as the most effective countermeasure, at 58 percent - the wrong part in the van is usually the result of an incomplete diagnosis on the phone.
When is a higher first-time fix rate not worth pursuing?
When the last percentage point could only be reached with larger van stocks or blanket parts replacement - then costs rise faster than the benefit, as Aberdeen noted back in 2013. Also when the rate is already above 85 percent and the remote resolution rate is the bigger lever, or when the company services special-purpose machinery whose cases rarely repeat.
9. Sources
All figures cited in the text with their origin. Studies by vendors with a commercial interest are marked as such. German-language sources are linked in the original - they are the actual evidence.
- Aquant, 2026 Field Service KPI Benchmark Report (press release of 19 February 2026; vendor study by a provider of AI service software; 161 service organisations, nearly 30 million service events, 7 million assets, 8.3 billion US dollars in service costs over three years): first-time fix rate, time to resolution, cost share of failed visits, remote potential, gap between technicians, workforce retention.
- Aquant, 2024 Field Service Benchmark Report (PDF, January 2024; vendor study; 145 service organisations, more than 24 million work orders, more than 582,000 technicians): measurement at 30 days, median by industry (industrial machinery 71.9 percent), 2.7 visits and 13 additional days per unresolved first visit, resolution cost 44 percent above the cost per work order. The 2025 edition (press release of 13 February 2025) supplied the figure of one third simple informational queries.
- kothes, KVD, KVA and SKDV, Insight-Report Service 2022 (PDF, November 2022; survey of service technicians in Germany, Austria and Switzerland between April and July 2022, 43.6 percent from mechanical and plant engineering, number of participants not stated; kothes is a technical documentation service provider): second trips caused by missing information, search times, own notes, paper, internet access on site. In German.
- Aberdeen Group, Fixing First-Time Fix: Repairing Field Service Efficiency to Enhance Customer Returns (Analyst Insight, March 2013, survey of 156 service organisations in January 2013; analyst firm; the PDF is hosted by a third party, Aberdeen itself no longer offers it): reasons for repeat visits, improvement strategies, 1.6 additional dispatches, customer retention by first-time fix band, share of organisations not measuring.
- TSIA, Top KPIs for Field Service Organizations (16 March 2021; membership association of the technology industry): median of 87 percent of on-site incidents resolved in one visit.
- Kundendienst-Verband Deutschland (KVD), Service Congress 2026 programme (24 August 2026): first-time fix rate as a steering metric in technical service. In German.
The first-time fix rate is the metric on which a service organisation can least deceive itself - provided it is measured with a fixed window, grouped tickets and a cause code. Whoever does that for two months knows whether the repeat visits are down to parts, experience or information. And whoever knows that no longer needs a vendor slide for the decision about the next step - only their own table.
