Summary

There are a lot of companies that specialize in selling to enterprises. How did they get their first customer? How did they know what to build? I don't have good answers for this but I did enough of it to recognize certain patterns. I've sent these notes to enough portfolio companies and friends doing sales to think they may be useful for the internet.

Here is a summary of the points:

• Enterprises take forever to close on deal, so you can't wait on them to start your company
• Once you close, they also take forever to integrate and deploy the product. It takes even more time to get useful feedback.
• When selling to enterprises you might go from 0 → 1M very quickly. Your underlying costs also grow quickly in unexpected ways after you've signed a contract with pricing. Consider that while drafting your first contracts.
• You might want to sell an enterprise a product but maybe they want to buy a service. I am defining a "product" as something that you hope to sell elsewhere in a repeatable fashion and a "service" as the direct work of smart people that will always need to be involved for the client to be happy.
• Enterprises pay for big contracts and they expect hand holding at all times. Consider how you will provide tailored support while the contract is ongoing. They will expect support regardless of what the contract says.
• Enterprises are very opinionated about what they want (also see product vs service). As their business changes after you've delivered your product they'll expect ongoing customization work. Make sure your contract specifies how they can ask for changes, what is in scope, and what isn't.

With that, my general advice is to:

• Start with the smallest customer that you can sell to that is still part of the larger market you are going for.
• Build for them, feed from their feedback.
• In parallel start enterprise sales. It will take more than a year for many of them to close.
• Start many, many conversations. Few will close.
• Don't ever wait on any of them to do anything.

Lessons from selling to enterprises

Time to close

Enterprises don't feel a sense of urgency. For anything that move quickly the urgency has to be there already or you have to generate it. This extends to the decision of buying what you have to sell. When you approach a company, make sure you understand how long it will take to close them. So you want to sell to the enterprise? beautifully captures how selling to the enterprise feels like. Read that before reading the rest of this post.

One fun way to gauge a company’s agility is to count how many distinct people show up to the first few meetings:

• 1 - 2 → you might close them in a few weeks
• 2 - 3 → you might close them after months
• 3 - 4 → you might close them after many months, maybe a year
• 5 - 6 → it will take at least a year, maybe more

Other heuristics:

• Do you have an “internal champion” who thinks it is their job to buy this product from you and is super motivated to do it? If not, you haven’t even started.
• Is your product in their top 3 priorities for the quarter / year?
• How many employees do they have? Think logarithmically: closing a company that has more than 10k employees on an important contract can easily take a year and a half. Closing a startup with 10 employees that is interested in your product might take 3 weeks.
• Do they have a procurement department? If so, expect more delays.
• Are they in a highly regulated industry? Banking? Healthcare?
• Are you talking to the highest executive already? If not, expect more delays “while they align the stakeholder”
• How long do they take to respond? Do they ever respond to emails on the same day?
• How often is there somebody “they need” that is on a break? Or simply doesn’t have time right now?
• Do new names of people come up on every conversation? “We should ask Mary about this” “Yeah and Carl too”.

And that is closing, we haven’t even gotten to deployment, implementation, or integration.

After closing: integration and deployment

Products that depend on work from the buyer and charge per usage learn to add clauses to their contracts to prevent their buyer from “defaulting” on the contract simply from not using the product:

• The implementation should start within N months of the signing of the contract, otherwise the buyer will pay a fee of $X.
• There are special discounts for the buyer that only kick in if they integrate before X date.
• etc.

Companies that do a lot of enterprise sales often hire “Customer Success Managers” which are meant to project manage the customer’s employees into implementing the product.

Consider your costs at scale

What happens if you charge Amazon per transaction and promise to store all transactions forever in case they need them? Unless Amazon grows exponentially (literally), at some point the ongoing storage cost of past transactions is higher than the revenue from new transactions. At that point you’ll have to break the contract or go bankrupt.

Beware the hidden consulting

Enterprises are used to getting a lot of customization from the products they buy, and for a good reason: they have a lot of users, opinions on how things should work, and they are paying a lot. As such, they will ask you to do custom work for them.

But there is a fine line between customizing your product (something you can sell repeatedly) and doing consulting work for them (something that only they need). People at enterprises hire consultants for a reason: they know they are often better at getting things done than themselves. They might inadvertently asking you to do a consulting project that you will not be able to sell anywhere else. A little bit of this is fine if you close big clients but make sure you understand what is going on.

They think they know what they want

While enterprise buyers insist that they need feature X, they are often wrong. Always check if why they want something. "Oh, because Accounting needs it for book closing.", ok, in that case, make sure you ask Bob from Accounting if that is the case. Accounting might need X, or they may need something else, or nothing at all.

Don't sell more than what you can chew

As you are selling "a platform for X", they might respond that they'd like that but only if it had a "solution for Y", and now that they think about it, then they also need a "system for Z". Eager to close a contract, you commit to X, Y, Z, and something else they asked for at the last minute.

Even if you wanted to add X, Y, and Z to the product, will you manage to do so before the customer is angry? I've certainly made this mistake, derailing entire engineering teams. It would've been better to tell the customer "We can only do X and Y, and that will take 6 months. Do you still want it?" and deal with a No if that were their answer.

Consider support and later customization costs

Big enterprises expect a high level of service. They want a person to talk to about their problems; preferably, always the same one. In the beginning, that is usually the founder. But eventually, you need account managers for large accounts and that is an ongoing cost.

Similarly, as their needs evolve, they will ask for further customization on the product. Different companies can be more or less reasonable here. But do put in the contract the extents of the initial customization and the way for them to get further customization.

Why chase enterprises?

But all this waiting is worth it. Companies follow a power law: the top companies are orders of magnitude larger than the smaller ones. Securing one big contract with Uber is better than selling to all the independent chauffeur’s in the world.

So, do start on the enterprise sales right away. But also understand that your new product or company can’t depend on them to close quickly. If so, you’ll die before way before they make up their mind.

Thank you to Tara Seshan for discussing the ideas in this post.

« Twitter's Sith and Jedi Pricing APIs »