eDiscovery Cost Estimate Template

Caldera Fluid Systems v. Ostrander Pumps  ·  responding party

Rate Card and Unit Map

The unit column is the point of the sheet. A card that lists prices without saying they are billed in eleven different denominators is why estimates go wrong in the same direction every time.

The two rows nobody asks for, and the one that stays empty

The last two rows are why the risk column can exist at all. They are what collection, processing and review cost when the work runs against a court-ordered deadline, and no vendor volunteers them.

Translation says not quoted rather than carrying an average. Two custodians correspond in German, nobody has asked a vendor, and a placeholder in a numeric column becomes a fact the first time somebody quotes the total.

Cost Model

Composition first, because every per-document line below is multiplied by what comes out of it.

379,240 documents of difference, which is $424,749 on the review line. The 4,772 is an output of the four rows above it, never an input.

Why 31 months and not 14

The same model run to the close of fact discovery totals $1,339,845. It is $72,318 lower and it is wrong, because the data does not come out of the platform when fact discovery closes. It sits there through summary judgment and through trial, billed every month.

Review is 81 percent of this sheet. It moves with the composition block at the top, not with the gigabytes, which is why that block is above it rather than in an appendix.

Scenario Comparison

Each option rebuilt from the top rather than scaled, because the phases move in different units and at different rates.

Then the half that is normally missing

Project management is the row worth staring at. It does not fall, because the case runs as long as it runs whatever the scope is, and saying so up front is what keeps the rest of the sheet credible when somebody checks it.

The penalty for being wrong on B is $181,583, which is 12.9 percent over doing the full scope once. On C it is $256,289. Both are smaller than people assume when they refuse to narrow at all.

Actual Against Budget, month 9

Read by driver, never by running total. Every variance is volume, rate, duration or scope.

Six percent under and twenty percent over, on the same day

The underspend is the largest line running nine weeks behind, which is not cheaper review. It is the same money spent later, alongside more months of hosting and a shorter runway that may attract the compressed premium.

The overspend is a 235 GB journaling archive nobody had assessed, a privilege rate running at 0.9 percent against the 0.6 percent modelled, and a motion that the Scenario Comparison priced exactly and the memo total left out.

Volume and scope move the estimate to completion. Rate and duration move the timing. A status update that reports the running total and not the driver is how a matter arrives at the end 20 percent over having looked fine all year.