We use Workable to manage our hiring practices and it's a great product. It does everything we need, has great UX and UI, and was simple to integrate into our processes (really, it helped our ad hoc process become a lot more structured in a good way). I logged on today to post a new job (BTW: we're hiring now and we have some more positions that we'll be posting for soon) and I realized we probably needed to move up to the next pricing tier. Then I got confused.
I didn't get confused by the features offered mind you, I got confused by the pricing.
When we started Breadcrumb, Seth Harris (the CEO of Breadcrumb) and I spent a lot of time working on the pricing model. We realized that charging for extra features would mean a compromised experience for those who didn't purchase those features and charging for users would place a barrier to adoption for venues with high volumes of staff (think: bars with different bar tenders working on different days), but charging per device would scale with the size of an establishment organically (larger venues require more POS terminals). We also wanted - as most pricing models do - to create an increase in value as a customer moved up the pricing chain. The pricing model we settled on was pretty similar to the pricing model that Breadcrumb Pro maintains today:
That makes for pretty clean curves if we map the pricing by tier. The absolute price moves up linearly and the per-device price falls off as you move through Tier 2 (Figure A). This reflects pretty conventional pricing wisdom: volume discounting. But when you look a little closer, there's actually a stranger curve at work. If you dial in on the per-device pricing, rather than the per-tier pricing, you see that the per-device price climbs at each tier initially and then falls as you approach the maximum value of each tier (Figure B).
What Figure B reveals is that the price per device actually goes up slightly as you pass through tiers 1, 2, and 3 but as you continue to add devices the per-device price drops. To be specific, when you go from 1 device to 2, each device becomes $0.50 more expensive, and $0.17 more again when you add that third device, but when you add the fourth you now save $24.92 per device - a serious value improvement - which grows when you go to device #5. Then, when you move from the $299 Tier 3 to the $399 Tier 4, that sixth device makes every device $6.70 more expensive than they were when you only had five however you begin saving again from the 7th device onward. In each case where the price went up it did so almost inconsequentially (<1%) or modestly but only for a brief increment (device 6).
So in Figure A, the green line represents the minimum price per tier based on the maximum number of devices possible in that tier. The marketing coaxes you into viewing the value proposition in its best light - when your brain sees $299 for 5 you decide that Tier 3 represents a per-device cost of $60 vs. the $99 of Tier 1 when, in fact, that $60 ($59.80 to be precise) is only realized when you have 5 devices at the end of Tier 3, not when you have 3 devices at the beginning. That said, for the reasons noted above, the majority of each tier is a significant value improvement over the previous tier and thus there is a perceived and realized benefit as you move up the tiers.
When we worked on our pricing model we thought a lot about this conundrum. What line do you want to strive for in the chart? If we wanted to ensure that value was attained immediately as you moved Tiers then we would have to create a pricing chart that was less immediately coherent (that nice linear line up in Figure A). If one device cost $99 then two devices would have to cost $98 each to be an improvement which would mean Tier 2 would be $196. That would not feel right when you looked at the pricing matrix. $196? That's a weird number to price this product at. It would create more confusion than the value it delivers. It's an unusual enough number that it would make you think about the pricing to try to understand the mechanics of why we would price something at $196.
The balance we struck - which I believe is both the most common and the most rational - was to maintain a coherent pricing increment through the tiers and an easy to comprehend value improvement based on the maximum number of devices in each tier. This had potential downsides for us (if it was the same cost for a user to have 6 devices as it was to have 10 they would be more likely to have a few more which would mean more server costs for us) but in the end, simple coherence won out over absolute pricing coherence and potential incurred costs for our business.
So let's get back to the Workable pricing model:
The initial reaction is nice. We see that clean - in this case slightly curving - progression of costs through the tiers (19/49/99/199/399). It's not a straight line but it's completely coherent to the eye and the brain (Figure C).
But when we look at the per-device numbers in Figure D things start to look really strange:Figure D (Those prices are hard to read so there is a data set to the right with the points.)
What you see in this chart is that the minimum per-tier price doesn't descend as you move up the tiers. It's $19 per job in Tier 1, then drops to $16.33, then rises to $19.80, then rises again to $19.90, and finally only drops in the top tier by virtue of it having an unlimited number of potential jobs, however it only becomes value-positive relative to the other tiers at 25 jobs or more. The minimum price per job is $16.33 when you have 3 jobs in Tier 2. That is the best price per job they offer in their pricing structure until you have 25 or more jobs. The difference between a company who has three job openings at a time and one who has 25 is massive and thus Workable has priced their product in a way that makes it less and less economical to use it the more jobs you have as a smaller (and growing) company.
"Noah, who cares?" you are asking yourself.
(I get that question a lot.)
Well, let's see what Workable is actually encouraging me to do by virtue of their pricing model. We currently have three job opportunities available, all of which are in our Australian office. We are about to post for a new job opportunity in our New York office. I figured I would just add another job in our current Workable account and use one account to manage both New York and Melbourne however that would be a mistake for me.
By going from three jobs to four I move from the Lite ($49) plan to the Standard ($99) plan and my cost per job goes from $16.33 to $24.75 per month while my annual actual cost goes (assuming I run the jobs for a year) from $588 to $1,188. Instead, I could start a new account and my cost per job is $17 (1 @ $19 + 3 @ $16.33), my annual costs go from $588 to $816 instead of $1,188 (saving $372), and it gets better as I grow. If I then want 3 jobs in the US and 3 in Melbourne, by using the two accounts I pay $16.33 per job vs. $33.17 per job in the Professional tier and my annual costs are $1,176 vs. $2,388 - a savings of $1,212 which is real money.
That buys a lot of nice things around the office to make our team happy.
I don't know if we'll do that but, because they made their pricing model incoherent, Workable made me think more about the per-unit pricing and that revealed a value problem which in turn revealed a way to get the same value for myself at a much cheaper rate. This potentially leaves a lot of cash on the table for Workable and generally affects my feelings about my relationship with them. When you make it more expensive to consume more of your product you make me feel that you're not a partner in my company's growth and then I feel less supported by you, less likely to evangelize for you, and less likely to stick with you in the long term.
All because of the pricing model.