Why Real Time Dashboards Matter More Than Monthly Reports

A monthly report tells a contractor what went wrong last month. Real time dashboards tell a contract manager what is going wrong on this shift. That sounds like a difference in presentation. It is really a question of how many days a problem runs before anybody can act on it.

This blog is a small research paper. One question. The published data that answers it. The finding. And what it means for a cleaning or security or FM contractor. Every figure comes from a named source or is calculated in a numbered table.

See this week and not last month. Operify AI gives contractors real time dashboards built on one live record where under delivery shows in hours and pounds by site while there is still time to fix it.

Key Findings#

• 25 days late on average. A monthly report issued 10 days after month end shows an event 25.2 days after it happens on average and up to 40.4 days. A weekly report shows it in 4.5 days.
• 14 hours lost before anyone looks. A site running 4 hours a week short loses an average of 14.4 hours before a monthly report shows the gap. A live record shows it the same day.
• 84 per cent past the RIDDOR deadline. If a monthly report were the first notice then 84 per cent of over seven day injuries would reach head office after the 15 day reporting deadline [1].
• KPIs are the weakest management habit. UK firms scored 0.42 out of 1 for KPI use in 2023 which was the lowest of the four practices the ONS measures [2].
• 5 to 6 per cent higher productivity. Large firms that make decisions from data have output and productivity 5 to 6 per cent higher than their other investments would predict [3].

1. The Question#

How many days does a problem on a cleaning contract run before a monthly report shows it and what does that delay cost against a live record?

Most contractors still run the business on a monthly pack. Hours from payroll. Scores from the quality audit sheet. Incidents from a log. Margin from the accountant. Each one arrives after the month has closed. Meanwhile the service plan promises hours every day and the SLA is judged on response times measured in hours.

The ONS gives a national benchmark for managing by numbers. In 2023 UK firms with 10 or more employees scored 0.42 out of 1 for the use of KPIs against 0.80 for continuous improvement [2]. KPI use was the lowest of the four practices measured.

2. The Data and the Method#

We combined six published sources with one illustrative estate.

Table 1. Sources used and what each one answers

The estate is the same one used in our last research blog. 300 operatives on 40 sites working 25 hours a week. It is a model and not a customer. We take a month as 30.4 days and assume a monthly report issued 10 days after month end and a weekly report issued the day after the week closes. Events fall evenly through each period.

3. What the Data Shows#

3.1 A Monthly Report Is 25 Days Late on Average#

An event on the first day of the month waits 40.4 days to be reported. An event on the last day waits 10. The average is 25.2 days. A weekly report cuts that to 4.5 days. A live record shows it on the day.

Table 2. Days from event to first sight and hours lost on a 4 hour a week gap

Modelled. Month of 30.4 days. The report dates are assumptions. The 4 hour gap is a site contracted for 180 hours a week and rostered for 176.

Four hours is small. Under delivery of that size never looks dramatic on a roster. Left alone for a year it is 208 hours of service the client paid for and did not get. The monthly report does not cause the gap. It decides how long the gap runs.

Scale makes it worse. The estate works about 32,600 hours a month. At the ONS absence rate of 2.9 per cent [6] that is about 945 absence hours a month. A monthly report shows how each one was covered between 10 and 40 days after it happened. The same delay hits every other line in the pack. A missed periodic. A toolbox talk that was never recorded. A competency matrix that went out of date in week two.

3.2 Some Deadlines Expire Before the Report Lands#

Statute does not wait for month end. HSE requires a report within 10 days for a specified injury and within 15 days for an injury that keeps a worker off for more than seven days [1].

Table 3. Incidents first seen on or after the RIDDOR deadline

Modelled. Share of incidents spread evenly across the period that would first reach head office on or after the deadline if the report were the first notice.

A serious injury is usually phoned in. The risk sits in the incidents nobody phones in. The near miss on a night shift. The slip that turns into eight days off. A contractor who learns about those from a monthly pack is already late. Capturing the incident at the point of work and deciding RIDDOR status at capture removes the wait.

3.3 What the Research Says About Data and Feedback#

Brynjolfsson and colleagues studied 179 large publicly traded firms. Those that made decisions from data had output and productivity 5 to 6 per cent higher than their other investments would predict [3]. The study measured the use of data and not screens. It still sets the prize that real time dashboards aim at. Data only helps when it reaches the decision in time and a monthly pack reaches it after the month has gone.

Feedback also needs the right target. A meta analysis of 607 effect sizes found feedback improved performance on average but more than a third of interventions made it worse [4]. Feedback worked best when it pointed at the task and not the person. Real time dashboards that show the site and the shift and the missing hours point at the task. A league table of supervisors points at the person.

Design matters too. A review of dashboard research found dashboards help decisions when users can drill down from a summary to the detail behind it [5]. That is the design brief for real time dashboards on a contract. The detail behind each figure is the time and attendance record and the checkpoint scan that proves a shift happened.

A monthly pack tells you a site was short. It cannot tell you which shift. Operify AI builds real time dashboards where every figure traces back to the bid line and the punch and the task and the audit behind it.

4. What It Means for a Contractor#

The test is simple. For each number in the monthly pack ask how many days old it is when somebody reads it.

Table 4. The same measure in a monthly pack and on a live record

The final column describes Operify AI as set out on its website. It makes no claim about any other product.

Five things separate a live record from a prettier monthly pack.

• Show hours before scores. An audit score means little until you know how many hours were delivered in that area.
• Point at the task and not the person. Site and shift and missing hours rather than a supervisor league table [4].
• Trace every figure to its source. A number nobody can drill into gets argued with and then ignored [5].
• Raise the problem unasked. A certificate that expires in seven days should arrive before somebody thinks to look.
• Turn a gap into an action. A failed audit becomes a reactive task with an owner and a deadline.

In Operify AI the Living AI module watches the record and surfaces the contract drifting into loss and the site whose scores are sliding and the certificate about to expire. Nobody has to open a dashboard to find them. The margin leakage view shows contracted against delivered hours and the margin on each contract this week and not next month. Both sit on one record so the cost per hour worked in the bid is the figure delivery is measured against. Agreed variations sit on the same record so scope creep shows as unpriced hours instead of vanishing into the month.

5. What This Does Not Show#

• The lags are modelled. A 10 day close is an assumption. A slower close makes every monthly figure worse.
• The 4 hour gap is illustrative. It is not a measured rate of under delivery.
• The RIDDOR shares assume the report is the first notice. Most serious incidents are reported directly.
• The productivity study shows an association. It does not prove that dashboards cause the gain [3].
• The ONS benchmark covers firms with 10 or more employees. Smaller cleaning firms sit outside it [2].

6. Conclusion#

A monthly report is a good record of the past and a poor tool for running a contract. On our model it shows a problem 25 days after it starts and lets a 4 hour gap cost 14 hours before anyone sees it. For incidents nobody phones in it lands after the RIDDOR deadline in most cases.

Live data does not replace the monthly pack. It changes what the pack is for. The pack becomes the record. The dashboard becomes the place where the week is run.

Bring us one contract and the pack from last month. In thirty minutes we will show the same month on real time dashboards and the days each problem would have been seen sooner. Book a demo.

FAQ#

What is a real time dashboard for a cleaning contractor?#

A live view of hours and tasks and incidents built from records made at the point of work. It shows this shift rather than last month. The base layer is time and attendance captured on site.

How late is a monthly report?#

If it lands 10 days after month end it shows an event 25.2 days after it happened on average and up to 40.4 days.

Does a weekly report solve the problem?#

It cuts the average to 4.5 days. That is far better and still too late for a shift that is being missed today.

What does reporting lag cost?#

On a site running 4 hours a week short a monthly report lets about 14.4 hours go before anyone sees the gap. Over a year the same gap is 208 hours.

How does reporting lag affect RIDDOR?#

HSE deadlines are 10 days for specified injuries and 15 days for over seven day injuries. If a monthly report were the first notice 84 per cent of over seven day injuries would arrive after the deadline.

Do dashboards actually improve performance?#

A study of 179 large firms linked data driven decisions with 5 to 6 per cent higher output and productivity. Feedback research adds that the effect depends on pointing at the task rather than the person.

What should a contract dashboard show first?#

Hours delivered against the service plan. Every other number depends on it. The gap between the two is under delivery.

Why do audit scores need hours data?#

Because a low score at a site that received two thirds of its hours is a staffing problem and not a cleaning problem. Scoring against hours logged tells you which one it is.

Will a dashboard replace monthly reports?#

No. The monthly pack stays as the record for the board and the client. The dashboard is where the week is run.

How much does Operify AI cost?#

Pricing is built around operatives and sites and the modules taken rather than a per head licence so cost stays steady through turnover. The figure is given on the demo call. See pricing or email hello@operifyai.co.uk.

References#

[1] Health and Safety Executive. When do I need to report an incident. RIDDOR guidance. Deaths and specified injuries notified without delay with a report received within 10 days. Over seven day incapacitation reported within 15 days of the incident. Available at HSE when do I need to report an incident (accessed October 2026).

[2] Office for National Statistics. Management practices in the UK 2016 to 2023. Released 13 May 2024. Firms with 10 or more employees. Scores range from 0 to 1. KPI use scored 0.42 and continuous improvement 0.80 in 2023. Available at ONS management practices in the UK (accessed October 2026).

[3] Brynjolfsson E and Hitt L M and Kim H H. Strength in Numbers. How Does Data Driven Decisionmaking Affect Firm Performance. Working paper. 2011. 179 large publicly traded firms. Available at Strength in Numbers on SSRN (accessed October 2026).

[4] Kluger A N and DeNisi A. The Effects of Feedback Interventions on Performance. A Historical Review and a Meta Analysis and a Preliminary Feedback Intervention Theory. Psychological Bulletin 119(2) pages 254 to 284. 1996. 607 effect sizes and 23663 observations. Available at Kluger and DeNisi 1996 (accessed October 2026).

[5] Yigitbasioglu O M and Velcu O. A Review of Dashboards in Performance Management. Implications for Design and Research. International Journal of Accounting Information Systems 13(1) pages 41 to 59. 2012. Available at the RePEc record (accessed October 2026).

[6] Office for National Statistics. Sickness absence in the UK labour market 2023 and 2024. Released 4 June 2025. Elementary occupations 2.9 per cent against a UK average of 2.0 per cent in 2024. Available at ONS sickness absence in the UK labour market (accessed October 2026).

About Operify AI#

We sell operations software with live reporting. That is a commercial interest in this finding and readers should weigh it. Two things constrain it. Every source is named so any reader can check it and every modelled figure states its assumption. Section 5 states what the data does not show.

Operify AI is operated by Codedevza AI Ltd company number 16485057. Correspondence to hello@operifyai.co.uk. Book a demo at operifyai.co.uk/demo.