The following is an e-mail exchange from 2013 recast as a Q&A with a start team launching a new service. Some identifying information has been altered. I have included an analysis at the end.
Startup: We have a complex pricing challenge for a new B2B service we launched 8 months ago that will be sold on a subscription basis to several distinct categories of business customer. We offer alerts to events based on keywords. Here is a chart we came up with to identify options, cost drivers, and concerns.
|Options for pricing/tiering:
- Number of alerts
- Number of entities followed
- Weighting of entities followed based on member count and/or total events they generate
- Number of users receiving alerts
- Setup cost for tracking a new entity
- Verifying feed for any event types an entity generates
- Operating costs are very low compared to setup
- It’s a new type of service in a new market.
- Possible customers range in size from individuals to small firms to large enterprises to non-profits to government agencies
- Plan to follow a wide range of events and entities
But we seem to be all over the map in internal discussions and cannot decide how to proceed. We have some grant money but need to transition to a business model to keep going.
Sean: I would start with a detailed analysis of the value to each of the different types of customers you have identified.
- What action will your customer take as a result of your notification or review of search results?
- What costs or risks do you help them reduce?
- What opportunities will you enable them to identify and take advantage of?
- What they are doing three minutes before they decide to add a keyword search request to your service?
- What result are they actually paying for? Is there valuing in knowing a keyword (or perhaps a synonym set) does not appear?
I think you have focused too much on what you can do and what your costs are, not enough on what customer is actually paying for.
Q: We agree, this is largely a value-based pricing question. We have many trial users that tested the service and originally said yes to certain pricing. But we are now covering more than 500 entities and several thousand events. Also we are selling a basic service now but believe we will have a long runway to increase the delivery of valuable data and insights over the lifetime of the relationship: getting relationships is key. Put another way, we believe we are in a “land grab” situation where getting the relationship is key and extracting value is really a secondary issue.
Sean: It’s a secondary issue until the grant money runs out. Also, there is a very big distinction between asking a free user to agree to certain pricing and actually getting them to pay. Either asking them if they will pay or how much they will pay typically has very little predictive value of conversation rate or price point validity.
Q: So if asking them if they will pay or how much they will pay does not work, what should we do instead?
Sean: Make them an offer and see if they do pay. If you are doing a “freemium model” now where you are giving away the service you need to either time limit access (e.g. a two week or 30 day or 60 day trial) or remove certain features and make them accessible only if they pay. In your situation it might be that some entities are free and the rest cost and/or some event types are free and others cost.
You should ask them what they are doing now and what it’s costing them in time, effort, dollars, opportunity cost, risk, etc… That’s factual. And from their answers you should be able to infer a value for your offering if it satisfies their needs and their constraints on a solution.
Q: We are still having trouble assigning a value (or a price), in part because we have several different customer types who we believe will gain very different value–or have very different ability or willingness to pay–from our service. We think the range might be as much as four times for the same subscription. We are not sure we want them to see each other’s pricing but are not sure how to hide it.
A: It sounds like your are more worried about leaving money on the table than getting people to pay. Normally your risk is that people don’t pay, not that you pick a price that’s too low and too many people sign up. It’s usually the case that there are other features the “high value” customers want and are willing to pay extra for. The trick is to engage and get them to pay something so that you can continue to work with them to refine and improve the service.
In addition to tiering pricing based on features or capacity customers are comfortable with different pricing for the same service based on other objective criteria: in particular enterprise customers accept that non-profits, schools, and students may pay less for the same service.
Q: We have an MVP, a good team that built our software, and a lot of interested customers but no pricing strategy. We have built a ton of models for various options and approaches but we are still struggling to go to the next step.
At some level setting a price, making an offer, and seeing of it’s accepted–or at least countered–trumps continuing to do more detailed analysis. As long as you have a theory for the value you are creating for your customer: your cost model is important to determining business viability but much less a factor in your pricing. You can make these offers individually or privately,
The negotiation can be more important than the opening offer especially for early customers where you are simply trying to move them from free to paid.
Here are some alternatives that are more in the nature of fixed configuration instead of user configurable you may want to consider:
- it may be easier for you to curate lists of keywords than to put the onus on the customer to define what keyword they want searched. This will also simplify your setup and testing.
- There may also be a value in selling a report with a set of common searches that address a certain fixed set of entities.
- There is probably a higher value letting the customer define a unique set of searches so that only they are notified, but less total revenue than a standard report.
- There may be value in indexing archives going back five or ten years to sell reports/briefings on trends and to provide a context for the last few times the same event was detected. This archive could also act as a training set to develop a taxonomy or ontology of key concepts.
Q: Thanks, we have decided to continue with our current open and free approach and negotiate individually with people who ask for new features.
Analysis: the fear of making a mistake can often paralyze a team. This seems to affect some teams more than others but all of us are affected to some degree. Early financing based on winning contests, grants, generous relatives, understanding spouses, or anything that does not come from a paying customer can de-focus a team from the need to develop a clear value proposition and business model. With money in the bank they can be distracted from making firm offers that may actually convert their “free users” into paying customers.
It’s OK to leave money on the table. In the beginning having a few customers pay something is an enormous risk reduction in your viability. Once you have established that at least a few people are wiling to pay something you can raise prices to where you actually make a profit on the offers that get accepted and achieve, or at least start moving toward, breakeven cash flow. It’s as important to plan for customer reference as much the cash value of the deal when you are getting started.
You can grandfather a small number of early customers with low pricing for a long time provided you are increasing prices for new customers and are on the path to breakeven cash flow.
It’s normally a good idea to be slow to raise prices on your early adopters: this recognizes the risks they took to embrace your offering early and the time they invested to help you refine your features and your sales and support processes. Instead of raising price ask for case studies, testimonials, and referrals once they are satisfied with our product. You can also negotiate on price in parallel with the level of reference and other terms and conditions.
You can continue to increase pricing over time as you add customers you can talk about, which reduces the perceived risk of your offering to new prospects, and as you move down the learning curve on onboarding and support so that can make firm promise sabout the impact of your offering on their business and how long it will take to achieve. Adding new features based on customer request and your deeper understanding of their needs also allows you to further differentiate and charge more, if only for a subset of your total customer base.