Ask three vendors what eDiscovery costs and you will get three numbers that cannot be compared. One quotes per gigabyte, one quotes per document, one quotes a monthly platform fee. None of them is lying. They are describing different parts of the same process.
This explains where the money actually goes, so you can read a quote properly.
General information, not legal advice.
The four cost centres
Collection. Getting data out of laptops, servers, mail systems and cloud platforms. Usually priced per custodian or per device. It is the smallest line on most matters and the one where cutting corners causes the most damage later — a defective collection is discovered at the worst possible moment.
Processing. Converting collected data into something reviewable: extracting text, de-duplicating, filtering by date and keyword, removing system files. Priced per gigabyte ingested. This is the number vendors quote when they say “cost per GB,” and it is the one that sounds cheapest.
Hosting. Keeping the reviewable set on a platform. Priced per gigabyte per month. This is the line that quietly dominates a long matter. A modest hosting rate on a large volume, running for two years, will often exceed everything else combined.
Review. Human beings reading documents. Priced hourly or per document. On most matters this is the largest single cost by a wide margin — frequently more than the other three added together.
Why “cost per gigabyte” is the wrong question
A gigabyte is a measure of storage, not of work. A gigabyte of scanned PDFs might be two thousand documents. A gigabyte of email might be twelve thousand. The second one costs far more to review, and review is where the money is.
So a low per-GB processing rate tells you very little. What matters is how many documents reach a human, and that is determined by what happens before review starts.
Where costs are actually controlled
Scope the collection. Collecting every custodian “to be safe” is the most expensive decision available. Every extra custodian carries through processing, hosting and review. Narrow the list with an actual analysis of who is likely to hold relevant material.
Cull before hosting. Date ranges, de-duplication, email threading and de-NISTing routinely remove a large share of a collected set before anyone pays to host or review it. This is the single highest-leverage step, and it happens early or not at all.
Prioritise the review. Ranking documents by likely relevance means the important material is seen first. If the matter settles or narrows, you have not paid to review the remainder.
Watch the hosting clock. Ask what happens to data when the matter closes. Hosting charges that continue after a case ends are common and rarely queried.
What to ask for in a quote
- Processing rate per GB, and what is included in it
- Hosting rate per GB per month, and the minimum term
- Expected culling percentage before hosting
- Review cost per document or per hour, and who is reviewing
- What happens to the data at the end of the matter
- Every fee that is not on the rate card — project management, productions, user licences
A quote that answers only the first question is not a quote. It is a headline.
The proportionality point
Federal Rule of Civil Procedure 26(b)(1) limits discovery to what is proportional to the needs of the case. That is not only a legal argument; it is a budgeting tool. If the projected discovery cost approaches the value of the dispute, that is a reason to narrow scope by agreement rather than to absorb it.
Smaller firms feel this most sharply. A process that is routine on a large matter can be uneconomic on a modest one, and the answer is usually earlier culling and tighter scoping rather than a cheaper vendor.
Related
- The Complete eDiscovery Process: A Guide for Law Firms
- AI Document Review: What Is Actually Defensible?
- eDiscovery & Litigation Support Services
Trying to budget a matter? Get in touch and we will give you a realistic range rather than a headline rate.