Master Service Agreement

Back when a contract was one document, every deal got its own. That single document carried the seat count and the price for this particular sale, and alongside them the rules about who pays how much when something goes wrong and which court hears a dispute.

The trouble surfaced on the second deal with the same customer. Clauses that had taken six weeks to agree the first time were sitting right there in the new document, and the customer's legal team read all of them again from the top. A new document had arrived, so procedurally it was a new review. Selling 20 additional seats took another six weeks.

Sales kept having the same exchange. "We already agreed all of this last time" was met with "that document and this one are different documents." Neither side was wrong. Rules that never change and values that change with every deal were written on one sheet, so touching one meant reopening all of it.

A Master Service Agreement (MSA) is that sheet split in two. The rules that rarely move get signed once, and the quantities and amounts that move every time go on an attachment. The first contract takes longer than it used to. Every deal after it closes on a single attached page.

What belongs here is whatever stays true across deals. The order varies by company, but these eight rarely go missing.

1. Scope — which services and which transactions the agreement governs. The sentence establishing that every later attachment hangs off this document lives here.
2. Payment — billing cycle, payment window, and interest on late payment, pinned to a date rule such as within 30 days of invoice.
3. Contract term — how long these rules stay in force. Separate from any individual deal's service period, and usually longer.
4. Auto-renew clause — the sentence saying the agreement continues on the same terms unless someone gives notice. The notice deadline written beside it is the real content.
5. Intellectual property — the product belongs to the seller, the data the customer uploads belongs to the customer. Without this line, whether that data can be used to improve the product becomes an argument later.
6. Confidentiality — often an absorbed version of the non-disclosure agreement (NDA) signed earlier, folded into this document.
7. Liability cap and indemnification — how much gets paid out when something breaks and who pays it. Capping at the last 12 months of fees is a common construction.
8. Termination and dispute resolution — for what reasons and with what notice either side can end it, and where a dispute goes.

If the service handles personal data, a data processing agreement attaches as a separate annex. It defines the scope of and responsibility for handling personal data on the customer's behalf, and because the law prescribes what it must contain, it hangs off the agreement rather than dissolving into it.

What's absent matters as much. Seat counts, prices, start dates and delivery schedules are not in this document. Those go on attachments such as a statement of work or a purchase order, so that one more deal means one more page.

The standard template usually comes from the selling company's legal counsel, and the person who carries it into the negotiation is a salesperson. When the buyer is a large enterprise the order reverses: their procurement team puts forward their own template and the seller marks up that.

Four groups read it.

1. Customer legal — reads the liability cap and the termination clause. Getting stuck here costs weeks.
2. Customer procurement — reads payment terms and renewal mechanics, hunting for any line that conflicts with internal policy.
3. Customer security — reads where the data is stored and how many hours pass before a breach is reported.
4. The seller's salesperson and founder — decide how far to concede. Anything the company couldn't survive, such as opening the liability cap to unlimited, goes to the founder directly.

The exchange happens through marked-up documents. The other side flags sentences it wants changed; you accept them or propose different wording and send it back. Three or four rounds is normal, and a first agreement runs four to eight weeks. The way to shorten it is preparation: send the template early and attach answers to the questions that always come. The three the other side's lawyers reliably ask are the liability cap, which country the data sits in, and the termination notice period.

"We signed the MSA last year, so this deal only needs the attachment."

The rules are already agreed; only quantities and amounts have to be written down.

"Their legal marked up the liability cap."

They won't take that clause as written, which means another round of negotiation.

"The MSA is signed but no purchase order has come through, so there's no revenue yet."

The rules are settled, but without a document saying they're actually buying, nothing has started.

At Pagedeck, Sofia once spent five weeks on this document with a 100-person customer. There was one sticking point. The customer wanted the seller to cover the full cost of any incident, and Grace wanted to hold the template's cap at the last 12 months of fees. On a $30,000 annual contract, unlimited liability means one incident can take the company down.

They settled by doubling the cap and putting data breaches alone outside it. The two later deals that added seats with the same customer closed on one quote and one purchase order. Those first five weeks bought the two days that followed.

Reading a signature as revenue is the most common mistake. It's called a contract and both sides signed it, which makes the assumption natural. But there are no amounts in this document. How much, starting when, is written on the attachment, and until that exists the deal hasn't started. Count signatures as closed deals in the forecast and the quarter ends with a hole in it.

Treating the template as non-negotiable is the other one. The company drafted it, so it ought to go through as written. In practice, the more purchasing process a company has, the more it rewrites sentences to match its own standards. Hold the template as a hard line and the negotiation stalls, and the customer's own champion loses the argument internally. Deciding in advance which clauses can move is faster.