Ugh, Pricing

Can price setting be scientific?

One of the problems with inventing a product is that you also have to invent its price. 

After a career of price setting, it never seems to get easy. Probably because as much as I’ve wanted to make it a purely logical process, it often comes down to a gut decision. 

All the Product Management sophistry advises the use of MVPs, iterating and test & learn methods. But customer’s reactions to shifting prices aren’t so forgiving. In SaaS (software as a service) prices can evolve gradually, but for most SaaS products, truly iterating on a price is a no-no.

As I approach launching HyThere’s first product for remote work teams, I’m reflecting on my past pricing decisions. Is there anything I’ve learned that I can apply this go-round?

It’s been awhile since I set a SaaS product’s price. In the early 2000s Capitol Advantage had a product for non-profits and trade associations called CapWiz. We were way underpriced. It caused all sorts of problems because clients didn’t see us as strategic. We also didn’t have the product margins to hire the engineering talent we needed.  We had product-market fit, but we didn’t have price-market fit.  

After about a year, I decided to raise prices. Not slightly, but I more than doubled them!

It was bumpy. We lost some clients, but not as many as I feared. And then good things followed. Our client meetings and sales calls began to be attended by executives. These leaders appreciated and valued our vision and wanted to do more with us! Our cash flow also improved which meant we could finally hire the engineering talent we needed. 

But there was one eye-popping change that I didn’t anticipate. 

The higher prices meant I could pay higher commissions. As a result our sales team crushed it. The same sales team with the same product at double the price more than doubled revenue. My pricing intuition had paid off. While I didn’t have the language for it at the time, I now know we were at an inelastic point on the price-demand curve.

It was a few years later at Custom Ink, where I began to develop my vocabulary around pricing theory. As it is, setting the price for any custom product is complex. There are dozens of factors that can affect the underlying costs substantially. Plus, in the case of custom t-shirts you have to adjust for whether the customer is ordering one shirt or 10,000 of them. Yikes! The permutations were mind-boggling.

Yet we were ambitious. Our team aimed to make custom t-shirt pricing entirely fact-based and scientific. We built an elaborate platform for testing prices and shaping the price curve. We embedded these capabilities deep into our systems and forecast models.  We doggedly pursued the elusive promise and potential for the perfect pricing algorithm.  

We learned a ton and had a big impact.

Yet despite operating the world’s preeminent custom t-shirt pricing data set. We still had to answer the most fundamental question with a judgment call, a hunch, an intuition... The question: What are we maximizing for? Do we want to optimize for revenue, gross margin, new customer acquisition, lifetime value or customer satisfaction? They were each possible… independently. But when maximizing for one, were we comfortable with the downsides for the others? 

Anyone familiar with machine learning recognizes this issue. The machines can’t do all the work for us. In the end every model still requires a human to make a value judgment. As much as we wanted to make pricing a science, it was still half art.

Target presented a different kind of pricing challenge. In big box retail, pricing is largely in the domain of merchandising, not the ecommerce team. Discount retailers must choose between a (mostly) everyday low price strategy or a (mostly) promotional route. Either way consumers ultimately get about the same price. That’s because everyone sells the same stuff and the internet makes prices transparent. Most items are ultimately priced similarly across retailers.

There was a twist when we bought Shipt in 2017. Like Instacart, they’re the gig-economy last mile delivery service. In this case, the pricing strategy for deliveries fell in the realm of our ecommerce team. Shipt was exclusively membership-based so people paid an annual fee. Target being the inclusive company that it is, didn’t want to require a membership with a big annual payment. Pay-per-delivery made more sense. 

But there was one massive wrinkle. Tipping. 

It’s no secret that the gig economy relies on tips. However people tip less, much less, when they just paid a delivery fee. But if drivers don’t make enough on tips, they won’t run the deliveries. The dilemma: how much could we charge for each delivery without overly impairing tipping?

I wish I could say we had a team of behavioral psychologists and economist PhDs working to optimize the delivery fee-tip mix. We did some lightweight testing, but nothing approaching Custom Ink’s sophistication. In the end we basically invented a price based on our gut. I’m not sure if it’s been tested since. I guess it was good enough.

Finally, there’s prescription drugs. I didn’t have much to do with setting OptumRx’s drug prices, but I learned a lot about how complicated it is. You’d imagine it’s simple. It’s a pill. It often costs less than a penny to manufacture. It should be the pill’s cost plus a pharmacist fee plus a profit margin, right?

Unfortunately, when it comes to drug pricing, the entire Rx industry has painted itself into a corner. For reasons I don’t think anyone fully understands, the industry uses the Rube Goldberg method of pricing. (Unfamiliar? My favorite example is “pass the salt”.) 

There’s no reason to dwell on Rx pricing here. But here’s my personal finance advice. Every time you fill a prescription, check both your insurance benefit price AND a discount app’s price. GoodRx is the most well known among a dozen options. It’s a total hassle, I know. But it’s the only way to guarantee a decent price. I do it every time and it saves real money. In fact, I just saved $40 going “off-benefit” last week.

I’ve set prices for all sorts of products in all kinds of industries, but here I am again: faced with inventing another price. It’s likely that I’ll go the per user per month route. But at what price points?

Now let me turn it to you. I’ve shared my stories. If you’ve recently built and priced a B2B software product, can we chat?  I’d welcome hearing your stories and learning from you!

P.S. A tactic I’ve yet to use is rebates. Dogbert has a shrewd approach.