Metabase vs Tableau for CPA Firms: Realization and WIP

Your practice management system, whether that is CCH, Thomson Reuters, or Karbon, already tracks billed hours, standard rates, and actual invoiced amounts. The first build should join those three into a single realization-by-client, realization-by-partner view, and pair it with a work-in-progress aging report showing how long unbilled time has been sitting. Metabase's SQL access makes this join straightforward once the underlying export or database connection exists, and a firm without a dedicated IT resource can usually get a working first version in a few days.

A useful first version includes these views:

• Realization by client, joining billed hours, standard rates, and actual invoiced amounts from your practice management system.
• Realization by partner, built from the same join, so write-down patterns are visible by owner.
• A work-in-progress aging report showing how long unbilled time has been sitting.
• A days-to-pay view by client, drawn from invoice payment dates, to surface late payers before cash flow tightens.
• A refresh at least weekly, so a partner sees an engagement running over budget in week three rather than at billing.

Between January and April, the number that matters most shifts from realization to raw capacity: how many returns are assigned per preparer, how many are still sitting untouched with the deadline closing in, and whether any preparer is carrying a load that risks quality slipping under pressure. A daily-refreshed Metabase dashboard filtered by preparer and return status turns a manager's gut sense of "we're behind" into an actual number they can act on, days before the deadline crunch rather than during it.

This same view earns its keep in reverse, too: it surfaces a preparer who is well ahead of pace and could absorb a return reassigned from someone falling behind, a rebalancing decision that is nearly impossible to make confidently from memory once a firm has more than a handful of preparers working simultaneously.

A single-partner or small firm can run all of this transparently, with every partner seeing every number. A larger firm, particularly one where partner compensation is tied to individual book profitability, usually needs each partner to see their own realization and WIP detail without seeing every other partner's client-level numbers. Tableau's row-level security handles that division cleanly. Recreating it in Metabase at that scale means real, ongoing permission management as staff and clients move between partners.

Disqualifier: skip Tableau if the firm's culture is fully transparent on partner economics already. The added governance is solving an access problem the firm has deliberately chosen not to have.

Write-down decisions delayed until year-end tend to be worse ones, made under time pressure rather than with a clear look at why a specific engagement ran over. Staffing cost is part of why this matters: median pay for accountants and auditors nationally runs $83,680 a year¹, so a senior preparer spending hours on an engagement that is quietly running over budget is real cost accumulating in real time, not an abstract number that only shows up at billing.

Once tax season ends, the same underlying data answers a different question: which advisory or bookkeeping engagements are worth expanding, and which clients are consistently unprofitable once realization is accounted for honestly. Many firms only look at this once a year, if at all, because rebuilding the analysis from scratch each time feels like too much work to justify. A dashboard that already exists and refreshes on its own removes that excuse, turning an annual scramble into a five-minute monthly check a partner can actually keep up with between deadlines.

The off-season is also the right time to revisit staffing decisions the dashboard surfaced under pressure during busy season but nobody had time to act on. A preparer who was consistently overloaded every deadline week, or a service line that never quite hits realization targets no matter who staffs it, is easier to address honestly in May than to notice again for the first time next February.