The Only #LitSupport Knowledge Base

Before I limited my law practice to work for courts and counsel, I was a trial lawyer working for contingent fees.  For 20+ years, I never charged for an hour of my time.  I funded the cases, did the work and was paid only if I recovered damages for my clients.  I charged 40% plus expenses; so, for the most part my clients and I shared roughly equally in the outcome.  At the time, I thought my fees proper, and they were certainly “industry standard.”  Everyone charged about the same, not from collusion but from plagiarism: lawyers didn’t draft fee agreements; we copied them.

But as I look back, I see that I could have charged less—even much less—and still have made a good living.  The only limits on what I could charge were the marketplace, where I saw no competition on price, and ethical precepts dictating a lawyer may not charge an illegal or unconscionable fee.

So, as someone who made a lot of money charging more than I needed to, I’m prompted to ask:

What are unconscionable charges in e-discovery?

I’m prompted by an affidavit, a sworn statement of an e-discovery expert leading a national litigation support vendor.  I won’t name names, because this post is about practices, not people.  But the affidavit shocked my conscience.

It’s no secret that e-discovery vendors can produce three projections of e-discovery costs:

1. The lowball projection that gets the business;
2. The accurate projection; and
3. The grossly inflated projection.

There’s a great deal of uncertainty attendant to projecting costs against unknown data volumes and variable data types.  Vendors need leeway to account for unpredictability in data volumes, data density and data complexity.  The projection that gets the business may embed the variables in dense contractual prose and present rosy, best case scenarios.  The projection of actual cost is the hardest to achieve, and ethical vendors labor mightily to supply reliable projections without losing business to less candid competitors.

But why would a vendor produce a cost projection that’s grossly inflated?

Four possibilities occur to me:

1. Vendor doesn’t want the business;
2. Vendor is incompetent;
3. Vendor thinks the lawyer and client are dumb or desperate enough to overpay; and/or
4. Vendor is pulling the wool over the eyes of a judge or master in an effort to curtail discovery.

Let me do the math, and you can comment on which of these (or other) possibilities adhere.

Vendor’s published rates are:

Vendor determines that data volume will be 5GB in “loose” data and documents and 500GB in e-mail messages to, by and between 25 co-workers and subordinates working on a common event/project that is the subject of the litigation.  So, the collected data volume is 505GB.

Some further assumptions:

1. Custodians extensively communicated with each other.
2. Communications were commonly sent to most or all of the custodians.
3. Attachments were common productivity file types and scanned documents.
4. Scanned documents would not undergo OCR as to do so would add $0.03 per page.
5. Collection costs are borne by in-house IT.  Data arrives ready for ingestion.
6. All volumes are uncompressed.
7. The projected volumes are based on solid metrics.
8. The court has ordered native production and banned TIFF conversion save for redaction.

The vendor projected costs as follows:

• Data Processing:

o   Deduplicate, De-NlST, Extract Text, Index Text = $27,775.00

o    Post-Processing Upload = $126,250.00

• Data Hosting

o   Hosting for 3 Months = $30,300.00

o   Monthly User Fees (2 reviewers) = $600.00

• Project Management (estimated 25 hours) = $4,375.00

TOTAL: $189,300.00

Did you do the math already?  Do you see the problem?

The vendor charges $27,775.00 (505GB x $55.00/GB) to process 505 GB of ordinary business data, virtually all of it e-mail.  This strikes me as high for work that is handled almost entirely by the machine.  Were I to process such data in my lab using, say, Nuix, I would expect it to de-dupe, de-NIST, extract text and index text in perhaps ten to twelve hours of largely unattended machine time.  I would also expect to achieve significant reductions in data volumes.  The reduction wouldn’t come from de-NISTing (which is wasted-but-harmless effort for this data) but from de-duplication.  Many of the e-mail messages and attachments will be held by all 25 custodians.  Most will be communications by and between pairs and groups of these custodians.  The loose documents will be frequently seen as attachments to the messages.  This is classically incestuous data.  It’s rife with repetition.

But, what volume does the vendor assume will emerge from de-duplication and de-NISTing?

Why, 505GB, the selfsame volume ingested!

The vendor then charges $126,250.00 (505GB x $250.00) to load the post-processed data set to a hosted review tool.  Just to ingest it…again.  It is, at most, a day’s work for one technician, with all of the heavy lifting done by the machine while the technician busies himself or herself with other billable work (or Netflix).  I grant that it may be a pass-through, marked-up cost from another vendor; but, that doesn’t make it any less a gross overcharge to those footing the bill.

The vendor then assumes that 505GB will be hosted for three months at $20/GB per month or $30,300.00, still using the overstated data volume.

$189,300.00 for this!  I think we have just defined “unconscionable charges.”

Are these projected charges 2, 3, 4, 5, 10 or more times what such services should cost if competently executed?  You tell me.  I charge hefty sums for my services; but, nothing like this.  If I charged like this, I’d fear losing my license for breach of professional responsibility.

And remember, this wasn’t just an opportunistic bid in a dog-eat-dog marketplace.  It was testimony, sworn to under oath as being true and correct on personal knowledge.

This was originally posted by Craig Ball on June 19, 2014 at https://ballinyourcourt.wordpress.com/2014/06/19/unconscionable/#more-1714