Can a SaaS pricing model change how people use your product?
Update: this story was originally posted on Medium. In the meanwhile I have switched to Ghost.org. Read the origin story on my blog.
I’ve been thinking about ways to change behaviors through pricing.
We’re building a CRM where we want our customers to build and nurture meaningful relationships with journalists. Less spam, more engagement.
Marketing and PR automation tools today are so easy to use and allow anyone, from anywhere in the world, to email a piece of content to a mailing list with a few simple clicks. That’s a good thing, right?
Yes. But if you want to build a good relationship, you need to keep that content relevant. I know, by now everyone is used to getting unsolicited email, and I can’t even remember the last time I read an outstanding newsletter in my mailbox. The problem is even worse for journalists, who get spammed every day.
In this post, I explore some concepts around how different CRM pricing models can change behavior for the better. Stick with me; I’d love to have your input.
Examples in the real world 🌎
There are plenty of real-world examples where price changes resulted in an increase of certain activity or have discouraged bad habits. Some examples:
- Increase the price of cigarettes -> less smoking (source)
- Carbon tax. Excellent Planet Money podcast episode about it
- Health-related food taxes (Fat, Sugar) change eating habits
Based on these links I am convinced that pricing can be used to affect behavior. Not only can it discourage bad choices (smoking, sugary foods, carbon emissions, …), it can also be used to motivate people to make better choices (buy fresh fruit, cycle, …).
In software
Can this principle be applied to software? Is it even possible to make marketing professionals send less spam and do more relevant outreach through the pricing model?
To find out, I searched for companies whose pricing patterns aim to encourage certain behaviors. I found a few:
- Annual vs Monthly. This is an obvious one where Annual plans are cheaper and will encourage prospects to commit to a longer term in exchange for a discount
- Multi-Factor Authentication discount. Mailchimp hands out a 10% discount if you set up multi-factor authentication
In the CRM space (Hubspot, Salesforce, Zoho, …), the pricing is what you’d expect: tiered plans (basic, pro, expert,…), user seat pricing and some subscription limitations (based on # of contacts, # of emails), or a combination thereof.
The exception I found was Convertkit, whose pricing is based solely on the number of subscribers (the size of your audience) — the more subscribers you have, the higher the monthly cost.
If we assume for a second that people take time to unsubscribe once they become uninterested, this is a good indicator of the success of your communication.
At the same time, if you send people a lot of crap, you will reduce your subscription cost (fewer subscribers). All in all, this pricing strategy assumes companies that send out more successful, targeted communications are likely to be doing more business, and so can afford to pay more, while smaller businesses and those that spam their contact lists are seen as less successful and so pay less.
What are we trying to do here?
In our case, we want to minimize spam, reduce the scale of large email campaigns, and encourage clients to do individual outreach and personalize their pitches. Trade the shotgun for a sniper rifle.
Model 1: Based on # stored contacts
This is an alternative pricing model a lot of CRMs are using. Clients are charged by the number of contacts they store and manage in their database. Easy to explain, simple to digest.
One issue with this approach is that people will work around these limits by not storing all their contacts in the database or, worse, game the system by deleting contacts to make room for another list right before launching a campaign, negating the whole idea behind having a “customer relationship management” system in the first place.
If there’s one thing you want a CRM to do, it’s to help people organize, manage and understand their audience. While this model encourages people to keep a smaller list with just their core audience, it still does not factor in the quality of the conversations had with that audience. You can still send 150 people a terrible newsletter every 2 days.
Model 2: Based on # used contacts
This should allow anyone to store their entire audience in a single database, which is one of the key benefits of any CRM, right?
In flagging contacts by “active in the last X days”, you can apply the same pricing scheme as in Model 1 simply by counting the number of contacts that are active.
Problem: Still no real indication of the quality and spamminess of your outreach. Sending 150 people a terrible newsletter every 2 days will not affect the monthly bill.
Model 3: Priced on # of emails sent
With this model, instead of looking at the number of contacts, we look at how many emails people send.
The major problem with this approach is that it completely ignores the quality of the outreach; in fact, it removes all moral guidance by giving you the freedom to spam as many contacts as you want, so long as you’re willing to pay for it. Capitalism at its finest.
One client sending one email campaign to 1,000 recipients with an open rate of 0.2% would pay the same as another client sending 10 monthly emails to a community of 100 recipients and scoring a 20% open rate. 😱
Not one of the common pricing strategies above helps push our goals of encouraging people to minimize spam, scale down large email campaigns and personalise their outreach. To change behavior I need another approach. How can we get the pricing to factor in the engagement of people receiving those emails?
Model 4: Baseline fee with penalties for spam
Keep a simple pricing model (different plans with price per user), but add extra charges for unsubscribes, bounces or spam reports.
This would discourage customers from sending 10k mailings to their entire database to announce that the CEO just bought a new Chihuahua.
Simple, effective. Yet I can’t help but feel like this approach uses the stick instead of the carrot, breeding ill-feeling and potentially driving away clients who — understandably — desire a more sympathetic, less punitive way of doing their jobs. After all, tools are intended to decrease stress and help you make informed decisions, not make you live in fear of tripping up and maxing out the team’s budget.
Additionally, people might see it as a dishonest tactic intended to maximize revenue. Heck, you could even argue that from that point onwards, we’d have a vested interest in encouraging our clients to spam people so that we could rake in more money — that is, if you’re looking at revenue as the all-encompassing goal of your business.
Model 5: Classify contacts in groups
Contacts would be classified in a few groups:
- Good: Contacts that engage with your content. Reply or click through to read more
- Unknown: People that are contacted but show no engagement whatsoever. It might be because of ad blockers or because they are on holiday. You don’t know
- Inactive: Contacts in the database that you did not try to engage with. This is a good thing, as we want to encourage people to send as relevant as possible. Sports journalists are not interested in the new version of your app.
- Bad: People that bounced, unsubscribed, etc…
- Terrible: Recipients reporting your email campaign as spam
Once that baseline is established, an example pricing model can be installed:
- Good: 0.2$ per contact
- Unknown: 0.5$ per contact
- Inactive: 0.0$ per contact
- Bad: 1$ per contact
- Terrible: 3$ per contact
Now, this model comes with a lot of challenges. First of all, it’s super hard to explain to prospects what their subscription price is going to be and how much it will eat into their budget, which makes it harder for them to pitch it internally. The reporting on the different categories needs to be transparent, where one can see the evolution of the size (and price) of each category over time; inevitably there will be arguments over what qualifies good contact engagement, bounced contacts and so on.
In short, there are a lot of moving parts here that while executable on an individual level, can make it difficult to scale this pricing strategy to thousands of clients, each with their own unique circumstances.
Model 6: Discount for high engagement
After publishing Marko Saric had a great idea:
Let’s say we apply a discount based on engagement (measuring engagement is a topic of its own). To oversimplify let’s say that engagement with content is either a reply OR click/read more click on your pitch.
To help me understand those numbers I quickly checked the average engagement rate of campaigns sent after January 2019. This is a data set of +17 million emails delivered.
Left axis is engagement rate where we’re counting clicks, opens, replies as engagement. Right axis is the average size of their campaigns. I grouped it per customer and kept it to the top 30.
We could cluster those results in 3 groups:
- Default: 0 to 5% engagement
- Good: 5 to 10% engagement
- Awesome: +10% engagement
One idea would be to offer people in the good cluster a 15% discount and awesome customers a 30% discount.
To ensure we keep raising the bar of the engagement rate we consider to be good or awesome we could keep this model dynamic and recalculate the average every month. Nice side effect: more accounts trying to get to awesome will drive the average up increasing the standard.
Conclusion
The first three concepts are simple and easy to understand and enforce, but will they change behavior? By writing all this down, I’ve come to believe there needs to be some kind of complexity in the dynamic pricing model to ensure that 1) it cannot be gamed and 2) it changes behavior.
Because of the complexity involved, such a model should be opt-in. This will allow clients to join the experiment and confirm that they too want to try to do the right thing — in our case, to reduce spam and improve personal outreach. To give that extra push, it needs to be clear that those opting-in will have the opportunity to save on their outreach — that is, if they behave.
Another big unknown in these alternative pricing structures is the client expectation: know-what-you-pay-for vs pay-how-you-use-it. How important is it for clients to know what their monthly bill is going to be? If every bit of budget spend has to be approved months in advance, would we be shooting ourselves in the foot by being unable to give clients a decent estimate?
These are all things we need to consider before diving into any great shift in something as fundamental as pricing structure. By solving this problem we would not only improve the quality of campaigns sent through our software, and thus our reputation in the market, we would also propel a fundamental change in how the PR profession operates and how it is perceived. But get it wrong, and we lose clients — and with them, the opportunity to help shape the industry to more sustainable and long-term thinking.
Do you have any other examples of companies implementing such a pricing scheme?
Share your questions, insights and ideas in the comments below.
📝 Save this story in Journal.
👩💻 Wake up every Sunday morning to the week’s most noteworthy stories in Tech waiting in your inbox. Read the Noteworthy in Tech newsletter.
