Recruitment Dashboards: What Should Agency Leaders Actually Be Tracking?

Most recruitment dashboards were built by exporting everything the customer relationship management (CRM) system could produce and arranging it on a screen. They are looked at during Monday's meeting and ignored for the rest of the week, because they answer no question anyone was actually asking.

Start with the decision, not the metric

Before choosing measures, write down the recurring decisions your leadership team makes. In most agencies the list is short and predictable:

• Which roles are at risk of not being filled this month, and where should effort move?
• Which consultants need support, and with what specifically?
• Are we going to hit the quarter, and what changes if we will not?
• Where is business development actually converting, and where is it wasted?
• Do we have the capacity to take on more work, or the headroom to hire?

Each decision implies a small number of measures. Build those, and only those. A dashboard grows organically from decisions; it becomes noise when it grows from available data.

Why the right metrics depend on your business model

There is no universal recruitment dashboard, and templates that claim to be one are the reason many implementations fail.

• Contract or temporary businesses live on contractors on assignment, gross margin per contractor, extensions, run-off and timesheet compliance. A placement count tells you little; the recurring margin book tells you nearly everything.
• Permanent businesses live on placements, average fee, time to fill and pipeline value weighted by stage.
• Retained or executive search works on assignment milestones, shortlist delivery dates and revenue recognised by stage: a weekly placement chart is close to meaningless.
• Volume or high-throughput desks care about funnel conversion rates and fill rates far more than individual pipeline detail.

A mixed agency needs the relevant view per division rather than one blended average, which flatters the weaker part of the business and disguises the stronger one.

Commercial: are we making money and will we continue to?

This is the leadership view, usually monthly and quarterly with a rolling twelve-month trend.

• Revenue or net fee income, against target, by division and by consultant.
• Placements or contractors on assignment, depending on model.
• Average fee or average margin, which explains revenue movement that placement counts hide.
• Pipeline value, weighted by stage rather than gross: an unweighted pipeline is a wish list.
• Forecast, with the assumptions visible. A forecast nobody can interrogate does not get trusted, and an untrusted forecast gets ignored.
• Client concentration, so you know how exposed the business is to losing one account.

Business development: is new work coming in?

Business development measures are the most frequently abused, because activity is easy to count. Track the sequence, not just the volume at the top.

• Target accounts and where each one sits in the relationship.
• Meaningful conversations, defined consistently across the business: a definition everyone agrees on is worth more than a precise number nobody trusts.
• Meetings held, and the conversion from meeting to job intake.
• Opportunities created, with value and expected timing.
• Job intake, which is the measure that actually predicts revenue eight to twelve weeks out, and job quality: how many intakes convert to placements, by client and by consultant.

Job intake against fill rate is one of the most useful pairs on any recruitment dashboard. High intake with low fill usually means the business is taking on work it cannot deliver, which costs more than having no work at all.

Delivery: can we fill what we have taken on?

This is the operational view consultants and managers use during the week.

• Active roles, by consultant and by client, with days open.
• Candidate pipeline per role, at each stage: the count that reveals whether a role is genuinely being worked.
• Interviews arranged and completed, and the interview-to-offer ratio.
• Offers made, accepted and rejected, with reasons captured. Rejection reasons are among the most commercially valuable data an agency collects and among the least reliably recorded.
• Placements, and start-date drop-outs, which are invisible on most dashboards and expensive.
• Time to shortlist and time to fill, which say more about your process than time-to-hire does about the client's.

Operational: is the process healthy?

Process health measures show problems while they are still fixable. This is the section most agencies are missing.

• Ageing roles: roles open beyond a threshold, particularly ones with no candidate activity in the last week.
• Stalled pipelines: candidates who have not moved stage in a defined period, which are usually a service failure rather than a data problem.
• Outstanding actions: overdue tasks, unreturned client feedback, references not chased.
• Compliance status: right-to-work checks, expiring documents, unsigned terms.
• Data quality: records missing salary, source or outcome. Poor data quality quietly invalidates every other chart on the dashboard.

The danger of measuring activity because you can

CRMs count calls, emails and logged tasks effortlessly, so those numbers end up on dashboards regardless of whether they inform anything. Three predictable consequences:

1. 1People manage to the measure. If calls are displayed, call counts rise and call quality falls. This is not cynicism about recruiters; it is what happens with any visible target.
2. 2Attention goes to the wrong signal. A consultant with low call volume and high conversion is a strength, not a problem, but an activity dashboard flags them.
3. 3Trust in the dashboard erodes. Once a team believes the numbers are for surveillance rather than support, data quality declines and the dashboard becomes actively misleading.

Creo's recommendation is to include activity measures only where they diagnose a specific outcome problem: for example, showing activity alongside conversion for a desk that is underperforming, rather than as a standing leaderboard.

Combining data from multiple systems

The reason most agencies never get a complete view is that the numbers live in different places: the CRM or applicant tracking system (ATS) holds pipeline and placements, the finance system holds invoiced revenue and margin, job boards and the website hold candidate source data, and the timesheet or payroll system holds contractor activity.

Bringing those together is a well-understood technical exercise where the systems offer an application programming interface (API) or a scheduled export. What makes it difficult in practice is definitional rather than technical: a 'placement' in the CRM and a placement in the finance system are frequently not the same event, and reconciling those definitions is the real work. Agree the definitions first, then build the pipeline.

• Agree one definition per measure, written down, before anything is built.
• Decide the refresh rate honestly. Delivery views often need to be live; commercial views are usually fine daily. Real-time everything is expensive and rarely used.
• Show the source and the last update time on the dashboard, so a number that looks wrong can be checked rather than argued about.
• Design for the device. Directors read dashboards on a phone between meetings; managers use them on a desktop while working a desk.

How many dashboards should an agency have?

Usually three, each with a named audience: a leadership view answering whether the business is on track, a manager view answering where the team needs help this week, and a consultant view answering what needs attention today. One dashboard serving all three ends up serving none, because the decisions are on completely different timescales.

Keep each to what fits on a single screen. If it needs scrolling, something on it is not supporting a decision.

A practical way to start

1. 1Write down the five decisions your leadership team makes repeatedly.
2. 2For each, name the two or three measures that would change the decision.
3. 3Check which systems hold those measures and whether the definitions agree.
4. 4Build the smallest version that answers those questions, even if it is manual for the first month.
5. 5Review after six weeks: which numbers were used, which were ignored, which prompted an action. Remove what nobody used.

That last step is the one agencies skip, and it is why dashboards accumulate charts. Removing an unused measure is as valuable as adding a good one.

Where the reporting problem turns out to be upstream (data not being captured consistently, or work not being recorded in the CRM at all), the answer is usually process and automation before visualisation. A dashboard built on unreliable data will be abandoned faster than no dashboard at all.