Is it time to do BI and where to start – step by step

Knowing where you are at and what’s important for you to measure

Professional services (PS) firms have a lot going on. There is a need to analyse all kinds of data, and the analysis is typically not straight forward. We typically see PS firms being interested in a broad range of KPIs, including:  

• Firm performance: revenue, gross profitability, work in progress, billing, capacity, target

• Firm efficiency: net efficiency , billable time, non-billable time, write-offs, timesheet coverage 

• Forecasting : pipeline (weighted/unweighted), deal landings, resources vs. unresourced revenue, deal flow dynamics  

Most firms pay attention to all of above, except the reporting is typically mixed between individual SaaS systems (e.g., Timesheets, Resourcing or Professional Services Automation (PSA) tools) and spreadsheets. It is not unusual for a firm to get its potential revenue number from a timesheet system and then to have a spreadsheet with a series of “hedges” and modifications applied to the revenue to get to the final number.  

Knowing which KPIs are important for you to measure is a good starting point. Understanding the complexity in getting to the final number you believe in for that KPI makes the decisions about BI much easier. For example, you might know that measuring revenue and capacity is time consuming and these are the two most important things for your firm to measure. You could take a view that those are the things that might be worth bringing into BI and leaving WIP and other KPIs to your finance system.    

How delayed is your management information?  

The first step is to understand which KPIs are important to you and what latency tolerance they have. If a KPI is important and its latency tolerance is low, it might be a candidate for BI.  

For example, job profitability KPI is highly time-sensitive. Consider a job spending £2k/day with a budget of £20k. It should take 10 business days to complete. Knowing on day 8 that you have run out of budget and putting controls in place (as well as having a conversation with the customer) is much easier than finding out on day 15 when you are already 50% over budget and can’t do anything about it.    

Such KPIs are candidates for business intelligence.  It would be incredibly difficult to maintain such information in a spreadsheet in near real-time. Additionally, spreadsheets are error-prone and have a limited capacity for rows of data, which means eventually you just run out of “space” in a spreadsheet.  

Who is the consumer of the management information?  

Understanding who might consume the management information (MI) and what types of decisions will be made as a result of this information will help determine accuracy of data you need, which ultimately governs how BI is approached.  

For example, if the management team only uses the MI to approximate overall firm capacity (something you might do in a spreadsheet) the calculation would be relatively simple and you could easily do it in a BI dashboard by combining data from your payroll system and people’s anticipated day rates.  This would make a good visualisation, comparing your revenue and capacity.  

On the other hand, if you are planning to set individual revenue targets for people, who will then see it on their dashboard, the data needs to be absolutely correct in real-time. You will need to consider individuals’ capacity based on their contracted work time, average write-off, holidays booked, holiday yet to be booked, average sick leave and a myriad of other parameters which would make the number exact. Such KPIs tend to be incredibly effective in managing people, but they rapidly lose their credibility if even the smallest error exists in the numbers.  A BI solution for this is a good choice, as getting a spreadsheet to do this is too complicated.  This does come with a health warning – if you are planning to build a bespoke BI (e.g., PowerBI or Tableau) the data engineering will take a lot of time, likely to be expensive to set up and to maintain.  

How available is the data?  

This may seem like an obvious point, but if your systems do not have good APIs, or APIs which do not export all the data, the BI effort will be futile.  This is less likely in today’s API-first world, but yet we still come across systems that won’t provide all their data on an API and leave only two options; replace or do it manually with spreadsheet.   

In Summary  

Most organisations today are looking at BI solutions for their business. It is a good idea, and they make a huge difference. The professional services space is quite specific, and its BI needs are typically complex due to the number of moving parts in the business (resourcing, jobs, revenue, billing, etc)  

We always recommend focusing on what moves the needle the most. Find out what takes you the most time, needs to be near real-time and has data available on an API. (This typically tends to be the revenue and the capacity.)  Once you have bedded in those KPIs move the business onto everything else.