As techies and nerds, we tend to believe in the utopians promise of technology enabling, at the very least, automation, and at the very best, intelligence. As consumers we have come to expect the “social contract” we have with the internet services we consume as arms-length “users” rather than as white-glove “customers.” As a result, as builders of magical web and tech products we come to view customers/users as an anonymous herd who must prefer to figure out what to buy and how much to pay on their own through endless screens of wizards, forms, and dropdowns instead talking to a human being. For consumer focused startups (e-commerce, social, gaming . . . etc) where lifetime monetization per users is at most in the hundreds of dollars, self-service transactions is an inescapable fact. But for entrepreneurs in enterprise markets, (saas, ad tech, b2b marketplace etc) we all too often believe self-service transaction and sales (usually in forms of some sort of purchase flow) is a development must-have rather than a feature to trade off. I’ve heard way too many entrepreneurs who believe the solution to building scalable sales channel is to not have one at all, but rely on potential customers to conduct the purchase with self-service via the website. More often than not, self-service transaction is a myth – a terrible joke and waste of time that many entrepreneurs and product managers come to realize only after spending months of engineering resources to find out no one cares.
Adoption. Self service is hard. More often that we think, customers or partners rather pick up the phone and have someone else to do the data entry for them. Take Google Adwords for example – after 15 years of being the only scaled source for quality traffic on the internet, less than 20% of its new customer revenues are from “self-service” – major brands and e-commerce/lead gen companies either uses agencies or have Google manage the campaigns for them. When startups that do not have the market leverage nor the history of value creation try to offer self-service, they must truly understand the transactional value of what they offer and the simplicity of the service they provide.
Price Discrimination and Value Pricing. Value pricing was one of the most important “innovations” of the late 90’s ERP boom. As it turns out (at least during that era), how much a customer is willing to pay for enterprise software is not how many engineers it took to develop it, nor the number of seats that the software is deployed to . . . but instead by the financial value/benefit the software can generate for the customer. Typically how this work is that a sales person will ask the potential customer a set of questions (product contribution margin, scale, sales process, cost structure, employee salary etc) and create an excel model to estimate the potential cost savings or even topline improvements the software can potentially bring if implemented. The sales person aims to charge between 10% to 50% of the value generated. The ironic artifact of this pricing scheme is that the sales process is more art and politics than it is science and math for a couple reasons. First, sales person is incented to go elephant hunting w/ the largest companies which typically have a much longer and convoluted sales process. Second, it is really hard to get customers to part with information that actually INCREASES the cost of their purchase – so it takes a really skilled and seasoned sales person. Using this technique, enterprise software companies can charge upwards of tens of millions of dollars per installation and generate tens of billions of dollars in revenue from simply targeting Global 5000 companies. It should be pretty much obvious by now that the process and outcome is pretty much impossible to replicate by building a bunch of web forms through self service.
Payment Method. A web self-service transactional funnel typically ends in a credit card based payment form. The other alternatives are not really optimized for self service – invoicing creates too much credit risk while ACH is not very secure for the enterprise applications. Given the typical limits on credit card transactions – the average transaction value is no more than $5000 and more likely in the hundreds of dollars range. Essentially a pricing ceiling is created when we forgo a sales person for a self service transactional funnel instead.
Sales Cycle. Enterprise sales cycles are measured in month – yet online web engagement is measured in minutes. Trying to force feed a “conversion funnel” into a month long decision making process is pretty much like threading a needle from a mile away. RFP type questions are hard to answer via a bunch of web pages – it is nearly impossible to credibly compare features, understanding pricing models, and understand technical limitations without a sales person hand holding the customer one step at a time via a series of meetings.
Decision Making. Enterprise decision making is not linear nor singular – but a website does make that assumption. That the user that is going through the self service transactional flow is the only decision maker and has the authority to pull the trigger on the purchase. This leaves a important part of the decision making process and their stakeholders up to chance where as a sales person can actually nurture and build consensus.
It is not to say that “self service” transaction will not work in all circumstances. In very very large horizontal markets where the actual number of potential customers is HUGE (in the tens of millions), forgoing the extraction of all the value of the product in exchange for sales and operational efficiency CAN work. Web analytics, web infrastructure, web site building, domain registration etc are all very large markets where there are just so many customers that a self service funnel actually CAN actually be a strategic advantage vis-à-vis a traditional go to market sales strategy. BUT only in very narrow set of B2B markets can billion dollar businesses be built on top of such low average transaction price. So chose your market carefully when trying to abandon the sales function for a self-service funnel.
