WhatClinic.com Blog

It seems to me that web application pricing has started to come of age.  I was getting sick to the back teeth of start-ups pricing on the 37signals model minus a few dollars. Just as the US lead the free model, now young eager start-ups in the states are driving realistic pricing that genuinely has the potential to create a successful company.

Over the next few weeks I’m going to have a look at a few of them and attempt to dissect their pricing with a view to seeing if their pricing passes the following tests:

1. Simplicity
2. Market Size
3. Comparable Pricing
4. Adoption
5. Price Discrimination

A Quick Introduction To KISSmetrics

KISSmetrics is an analytics service that attempts to model a website’s most commercially important interaction with its users – the conversion funnel. Their market is anyone who runs a website commercially and who cares about the performance of their conversion funnel. Value is delivered by providing business intelligence that enables the website owner to intelligently change their interactions with their users, maximizing their revenue as a result.

KISSmetrics Pricing

Test 1 – Simplicity

KISSmetrics has beautifully simple pricing that poses no barrier to comprehension. I can immediately see what pricing plan suits me best.  It almost seems a shame for them to include standard features on this list (Unlimited Reports, A/B Testing, GA Integration) which interfere somewhat with the simplicity.

I have one minor problem with the simplicity of the pricing – what happens if I opt for the 1 million event plan and in one particular month I run 1.1 million events?  Are the last 100K events trashed? Are they stored and subsequently computed if I upgrade. Is a random 100K chunk lost?  In my opinion this needs to be clarified and simplified in the customers’ eyes on the pricing page. For example, change the Events per month values to “First 1 million events”, “First 5 million events” and “First 10 million events”.

Test 2 – Market Size

Pricing affects market size in a variety of different ways. Trivially, market size equals the number of customers multiplied by the average price, so the higher your price the bigger the market. However, higher prices can reduce customer numbers and you can end up excluding part of the market. However, conversely, you can increase your sales/marketing resources allowing you to capture more of the market.

In my view there is no question that KISSmetrics have set pricing high enough so that even if only a small percentage of the market adopts their service the company will be a success. The pricing should also provide KISSmetrics with sufficient resources to be able to go after more market share.

I only offer one proviso here: if KISSmetrics is not as well funded as it seems, then it would pay to offer a heavy discount for annual subscriptions, rather taking slightly more money in the long term delivered as smaller monthly amounts now. That way the money invested in sales can be immediately recycled.

Test 3 – Comparable Pricing

No one wants to pay too much for a product. If a potential customer feels a service may be priced too high many will delay or avoid the purchase. Unfortunately for KISSmetrics the product that is immediately comparable to the first time visitor’s eye is Google analytics, which is free. To make matters worse, KISSmetrics itself ensures that I make this comparison by including ‘Google Analytics integration’ on its pricing page.

In my view, if you cannot provide a favourable external comparison then your own pricing needs to be comparable. I should be able to compare one of your prices against another and get a level of comfort that the package I’m buying is priced competitively. For this to happen two price points need to be close together. KISSmetrics’ pricing fails here.

Test 4 – Adoption

I don’t care what planet you live on but $1,800 a year is a lot of money.  Having entry level pricing at this rate automatically excludes a huge portion of the market that would otherwise derive value from the product. Unless KISSmetrics has high costs associated with a sale (maybe support costs) I would be tempted to look at a lower entry price that was limited in such a way that it wouldn’t cannibalize my higher priced offerings.

KISSmetrics Potential Market Size

This way it would be easier for KISSmetrics to get much wider adoption, and if the pricing was designed correctly they could push customers up their price plans as they prove the value of their product.

Even if a small user with 100K events per month was willing to pay $1,800, the pricing is going to make them feel like an idiot. “Here am I paying for a million events when all I’m going to use is 100,000”. Pricing should make the customer feel like the smartest guy in the world.

Test 5 – Price Discrimination

The goal of price discrimination is to segment the market according to willingness to pay, with a goal of maximising revenue.  KISSmetrics discriminates on one axis only: volume. It’s even debatable if they use price discrimination at all – one bar of chocolate costs X, two bars cost 2X, etc.

Since the distribution of websites by their traffic is a classic power curve (as in the events graph above) the KISSmetrics pricing model will always result in the most customers being on the lowest price plan.

This means that the bulk of the market can only give KISSmetrics $149 a month. Customers with less than 1 million events still get access to all of the features so there is no reason why they would ever adopt a higher price plan. This doesn’t make much sense; it’s kind of like offering a student price for a haircut and then not allowing them to purchase an expensive and more profitable colouring.

What defines willingness to pay for KISSmetrics? This is always difficult to model and generally we have to accept an inaccurate model that works for the bulk of customers but fails for a minority (for example cheap OAP pints fail to account for the millionaire OAPs). For a KISSmetrics’ customer it would seem that the factors that matter are:

• Volume of events
• Dollar value of  the margin on the average event
• Perceived potential improvement that can be expected

The second two factors are difficult to model which is probably why KISSmetric’s have stayed away from pricing off of them. However, I would contend that there are several ideas that would we could look at to help define these factors, albeit inaccurately.

• Logins. The number of people who want access to the data. A company selling X units at a high price is typically going to have more people who want access to business intelligence than a second company selling the same volume at a lower price.
• Accounts reconciliation. Refunds and charge backs are likely to be a factor in higher ticket items.  Adding a feature that would allow for the service to be reconciled with month end accounts ensures that refund sales and charge backs are taken out of the analytics and that data is true and accurate. In addition, top-end products may result in the creation of a sales lead rather than an online purchase and being able to reconcile sales with the lead conversion funnel will be valuable.
• Traffic Spike Overruns. This feature would securely store events that overrun a plan and can subsequently be recovered by paying the transaction fee. The longer a site has been in existence the more iterations it will have been through and given that KISSmetrics is targeting metric driven companies the less perceived improvements there are likely to be. It seems likely that older sites have more predicable traffic with smaller spikes,  so pricing off of this may be able to segment the market (needs more research).

Suggested New Pricing

I’m not going to suggest that this is right; however it should show the rough direction that I would like to move the pricing towards.  Also please accept that I haven’t put in the kind of effort required to make the verbiage easily consumable – clearly a lot of work would be required to get this into any kind of finished form.

Proposed New KISSmetrics Pricing

* All accounts get free A/B testing and commoditized website analytics integration (including Google Analytics).

This gives me

1. A much more attractive market entry price that is sufficiently limited so that it shouldn’t cannibalize my higher priced plans
2. Pricing is comparable. Silver is obviously much better than Bronze for just 20% more
3. There is an attempt to segment the top end of the market and to price discriminate accordingly

What do you think of KISSmetrics pricing model? How would you change it to move the company forward? Share your thoughts in the comments.

Also feel free to suggest other web application pricing that I should look at