One of the toughest things to get right in an early startup is the pricing. I often see technical founders price their products too cheap. They sometimes have a “cost plus” model in mind, applying a suitable markup to their costs to determine a price. However, for most software products, cost has very little bearing on value.
In the early days, sometimes you’re so aware of your product’s shortcomings that you don’t feel comfortable charging a lot. When you’re first getting started, it’s fine to leave some money on the table. The most important goal in the early days is to get to product / market fit. If a customer is paying you, that’s a great sign and you don’t need to optimize every deal. You can and should adjust your pricing over time to find the right price levels.
That said, many B2B startups under-price their product. They bring a small-business mindset to mid-market and enterprise customers. Small businesses are notoriously cheap and generally have the highest churn rates. Larger mid-market and enterprise customers are often more demanding. They have bigger problems but they also have much bigger budgets. You might consider offering different editions or plans that reflect the advanced needs (and budgets) of larger customers.
I talked to some founders recently who were planning on giving their product free to early “design partner” customers. I understand the spirit of this, because there can be a good exchange of information on both sides. However, when you give away a product for free, it’s not a real test of product/market fit. If buyer doesn’t have skin in the game, I don’t think you’re going to get their attention in the same way as when they are paying. I encouraged them to give those early customers a discount, but still to charge them. After all, there is no greater proof of value than when a customer pays you.
Beware Commodity Pricing
Startups sometimes try to use pricing as a way to disrupt a market. When used in conjunction with other elements, pricing can be part of a disruptive strategy. But on its own, commodity pricing is a mug’s game. It can easily lead in a race to the bottom with lousy margins for everyone. Competing on price attracts the most price sensitive customers and they don’t have a lot of loyalty.
Business buyers, especially in the Mid-market and Enterprise segments are far less price sensitive. They aren’t looking for the cheapest solution, they’re looking for something that will solve their problems and save them time or money. Serious buyers aren’t looking to cut corners in order to save the company a few bucks.
You still need to work within their budgets and understand the buying mechanics, who has authority, what approvals are required, etc.
It’s usually best to price your software in a manner that is familiar to buyers, whether it’s a seat-based model (like Zendesk or Duo Security), a server based model (like MySQL) or a consumption model (like AWS.) There’s plenty to communicate about your product and you don’t want to distract from that by introducing a complex pricing scheme. Generally speaking, simpler is better.
Even for software products that have a lot of complexity under the hood, it’s best not to foist that complexity onto buyers. Or said differently, the more time you spend explaining your pricing, the less time you have to talk about the benefits of what you’re selling.
Watch for Negative Margin Customers
If you’re unsure of what the cost will be of running your software once it lands in customers hands, then it might be best to plan around average use cases and ensure that in aggregate you are generating positive margins. If it turns out there are a small number of degenerate use cases where customers are costing you more than they bring in, you can either re-architect your software to reduce some of the underlying costs, re-negotiate with those customers, or put more aggressive limitations through your terms of service. What you don’t want to do is keep adding negative margin customers as you grow.
In the early days of Zendesk, we had a rough metric for thinking about the average “cost per ticket” based on storage and computation usage. Every now and then we’d have a one or two seat customer who was managing tens of thousands of tickets through an automated process. That meant they were costing us more than they were generating. We reached out to those customers and told them to either buy more seats or we would have to shut them down.
Use A Pricing Model Customers Understand
Products in related categories will determine acceptable pricing bands. Zendesk’s per-seat pricing model meant we could tell customers it worked the same as Salesforce. However, it also meant it was difficult to charge more than Salesforce did for their CRM. Luckily over time, Salesforce continued to find and creative ways to charge customers more and we were able to introduce higher priced editions with premium services.
Pricing your products lower than the competition can be a useful in positioning. That said, the pricing gap must be significant for it to work. If you’re 10 or 20% cheaper than the competition, no one cares. They can easily negotiate that with their existing vendor.
Can Pricing Be Disruptive?
We made a case at MySQL that it was 90% cheaper than Oracle. That said, we weren’t trying to migrate Oracle workloads to MySQL. There are few things harder to displace than a working database! If all goes well no one notices and if anything breaks, heads roll.
However, we were looking for new web uses cases where the complexity (and expense) of Oracle was not required and this was a way to encourage customers to look beyond their existing approved vendor list. Competing solely on price alone would not have been an effective strategy. Other vendors who pitched “Oracle compatible” databases never fared particularly well. It was the fact that MySQL was optimized for different workloads (web applications, scale out, embedded etc) that drove growth. Because it was freely available for download, MySQL snuck in the side door of many companies and got broad adoption internally before the CIO even knew about it.
However, Open Office, a free open source desktop alternative to Microsoft Office, never went beyond niche adoption from Linux fanatics and open source die-hards. Google Suite took a different approach, focusing on web users and built a multi-billion dollar business over time. In the early days, it wasn’t better than Microsoft Office by traditional criteria. It didn’t have the breadth of features that power users needed. But it was simpler and substantially better in one key area: online collaboration.
When In Doubt… Test
Pricing will evolve as your company grows. If you lose deals to competition based on price, that may require you to make adjustments. Sometimes, the biggest competition you face is your own low-priced “starter” level pricing. If you’re offering a low-end product on your web site for $19 per month per seat, it can feel like a bridge too far to get larger customers to pay $99 per seat for the Enterprise offering. The most important thing to do is to strengthen the differentiating features on your higher tiers.
You may wish use automated A/B testing on your web site to try out different pricing levels. Alternatively, you can use promotional campaigns to test changes in pricing, bundles or other variations. Pricing promotions work best when they are targeted to specific scenarios that map to the customer’s situation. When you’re getting ready to release your first product it might be a promotion for “early access customers” where you include a certain amount of hand-holding to ensure early customers are successful. Later on you might offer special migration pricing to woo customers from a competitor by including all of the integration tools and handholding to be successful.
If there’s a lot of complexity in pricing your product, you might consider a promotion that bundles all of the required capabilities into one “jump start” program, shielding the customer from having to worry bout all the underlying component costs and services.
How has your pricing evolved in your company? What experiments have you run that give you conviction about your current pricing?