My own experience as a founder was launching a self-serve model for SaaS, even though we eventually moved upstream to enterprise. Without thinking too hard about it, I just assumed that most businesses would do the same: start off trying to reach smaller companies whilst they got their product right and then move up to enterprise. But it’s only since working with other SaaS businesses that I’ve realised that this is often not the way it’s done.
This week I’m going to talk about the two approaches and what we can learn from each.
Self-serve
So let’s start with self-serve. What does that mean? It’s where customers can find out about a product, can even try it, and then when they’re ready, buy it and make it work for their needs - all without talking to a sales person (or at least very light-touch).
Companies such as Slack, Mailchimp, Trello, Notion, Intercom, SurveyMonkey etc all use a self-serve model.
Characteristics of self-serve SaaS
The thing about self-serve is that, to an extent, it has to be designed like consumer software. When you first set up Facebook, you didn’t expect someone at Facebook to set it up for you and you didn’t go and hire a consultant either. The product is designed to be intuitive for the user without the need for additional human intervention. It means that in most cases, with self-serve, great User Experience is baked in from the start.
Self-serve normally equates to lower Annual Contract Values (ACV). Because the sale is light-touch or no-touch, the product is normally less complex. Slack starts off as free and the pro version is $7.25/user/month. Whereas Salesforce Professional, more of an enterprise sell, is $80/user/month. If you are looking at lower ACVs, that means that your Customer Acquisition Costs (CAC) have to be keenly tracked. If your ACV is, say $1,000 and it costs you $3,000 to win each new client, that’s not a great metric (you want your payback period to be less than 12 months). For great self-serve SaaS businesses, this relentless focus on keeping the CAC down whilst driving growth, makes for some awesome marketing people who are under pressure to demonstrate an ROI for every dollar they spend. Firstly they will have the data because they are doing this at scale, but secondly they will be testing and testing again not only what drives leads and conversions, but also whether those conversions are good-fit customers (ie ones that don’t churn after a few months).
Self-serve SaaS tends to be more product-focused than its enterprise peers (more about that below). Yes UX has been well thought through, but the product itself has to achieve everything that is needed for the customer to use it. There are no options for humans to jump in and connect the dots where needed. A good self-serve SaaS has figured out what the majority of their customers need, to solve the job they are hiring the product to do. An off-the-shelf website builder isn’t going to ask you to ‘get in touch’ when you want to set up a custom domain, for example. You wouldn’t expect someone at a self-serve HR platform to send you an email asking you for a .csv file of all of your current employees for them to upload.
Self-serve tends to be scrappy, agile and entrepreneurial. In the early days of ScreenCloud we saw several trialists telling us that once we introduced portrait-orientation for their screens, they would sign up. We were desperate to get people using it and so we scrambled to make this a priority - sure enough, we increased our conversion rate virtually overnight. This immediate ‘change here: immediate results in uplift there’ is harder to observe in enterprise SaaS.
The big challenges for self-serve businesses moving upstream tend to be around building and scaling an enterprise sales team and articulating ‘enterprise’ value beyond simply better security and workflows.
Enterprise
For me enterprise is not only based on the size of the company you’re selling to, it’s also based on the type of solution you are offering and your Go-To-Market. Products like NetSuite or Dynamics 365 owned by Oracle and Microsoft respectively might jump to mind as enterprise solutions. But products such as the revenue management tool, Paddle, Data monetization platform, Zuora, would also qualify. Many SaaS companies are both (think Hubspot).
Characteristics of Enterprise
Enterprise SaaS companies have had to really graft to grow. That’s not to say that self-serve isn’t hard work, but you’re winning clients at scale, many of whom you never talk to. Enterprise by contrast requires you to speak to each and every one of your customers, often challenging conversations over an extended period of time. Really digging into the problem you’re solving for them and articulating is way more important than it perhaps is for the fast-moving world of self-serve.
And if you started off as an agency or consultancy, “enterprise” is more of a natural progression: you just start SaaSifying some of your processes and sell to the same type of customer. You go from delivering something 100% bespoke to say 40% bespoke with an underlying set of modules that you get customers to subscribe to. The fact that you’ve already spent your working life selling services to enterprise customers means this approach isn’t a huge leap.
And even non-agencies might find it easier to win larger amounts of money earlier on by selling higher value contracts that have a SaaS element with a substantial dose of professional services. It means you can go out with an unpolished product and get paid tens of thousands straight away, rather than wait for the self-serve experience of tens or hundreds to trickle in. But while this may be great for initial cashflow, it can cause problems.
I mentioned the importance of CAC above: for enterprise they may not be as on the ball when it comes to this key metric. If you are winning a handful of $50k contracts per year, how focused are you going to be on CAC? In my experience, not very much. I’ve seen SaaS companies (plural) doing well north of $1m in ARR who couldn’t tell you what their CAC was, much less their LTV:CAC. And to an extent you can understand it. Why waste time on working that out when there are massive deals to be won? And in any case, what are you going to do to bring your CAC down? Not attend that event? Not have a sales person in your customers’ territory? Fire your only marketing person? Of course not. The issues start when you want to start going for mid-market or SME/SMBs and your only known route to market has been expensive because up to now the ACV easily accommodates it. It also becomes very important if you are scaling your enterprise quickly. If the plan is to double your growth in the next 12 months, you’re going to have to more than double your sales and marketing costs in that time too to drive the leads within your sales cycle. Being more efficient with your CAC by being more aware of your ROI might have a big impact on overall capital efficiency.
I said above that self-serve SaaS tend to be product-focused. What I meant was that the product is more important than any one customer. It means building what’s best for the majority of your best-fit customers. With enterprise, the tendency or at least the danger is that products become individually customer-focused. I know of a company whose first customer generated $1m in ARR. Now, what degree of influence did this one customer have on the product roadmap? You can guess, right? What about the next customer who came along and didn’t need half the stuff that was in the current feature set?
In theory, any SaaS service could be automated. Something that is designed by hand could have a self-serve element with easy to customise templates. Data sat in one system could be synched to another without the need for a manual data dump. Something that needed a load of configuration could be done using AI. But if your SaaS has always come with human support then the cost and disruption that would cause based on the complexity and technical debt may make the prospect somewhat daunting.
All of this means that for enterprise SaaS businesses, moving downstream is harder. Reducing sales and marketing costs by moving towards self-serve is tough. Saying to customers that the product only does 90% of what they want is tough when you’re used to giving them 100% of what they ask for. And even finding people who know how to run a self-serve marketing engine can be tough, too.
What can we learn from each approach?
I guess you could conclude that if your plan was to have both an SME and Enterprise play you should probably start with self-serve and move up stream. Or you could just acquire a self-serve business. That’s what Salesforce did when they bought Slack. If that was the plan, I would definitely advise companies who don’t need to talk to every customer, to nevertheless have those conversations early on. Just because you are selling thousands of subscriptions, don’t assume you know the customers’ fundamental pain points. Take a leaf out of the enterprise guys’ playbooks and build a solid understanding of the personas and their Jobs To Be Done. This will give you better insights as you move up stream.
By contrast, and even if you never intend to be a self-serve business because you have no interest or your particular niche is only applicable to enterprise clients, there are questions to ask when we look at how our self-serve colleagues operate. For example:
Can we put as much care and attention into our product UX as a B2C SaaS would?
Even where a consultative sale and a level of professional services are required, is there still a case to lean towards self-serve wherever possible? To harmonise the codebase and have consistent documentation. To reduce the level of ‘touch’ that the customer needs from both sales and support.
Can we make bespoke features universal? Can we push to be multi-tenanted from the start, or move in that direction?
Can the onboarding and training be automated? Could it be in-product or delivered via AI?
Should we be as obsessive about metrics as a self-serve SaaS company would be?
None of this stuff is easy. But if self-serve SaaS companies can adopt the enterprise's proficiency in understanding and articulating the real problems they are solving for their customers, and enterprise companies can integrate some of the SME SaaS practices related to automation, user experience, metrics, and rapid testing, it will make life better for both in the long run.