Did you hear the one about SaaS application response time? Stop me if you’ve heard it before-
SaaS Application #1: How do you keep a user in suspense?
SaaS Application #2: I don’t know, how?
SaaS Application #1: …
SaaS Application #1: …
SaaS Application #1: …
SaaS Application #1: server timeout
Actually, just like that Dad Joke adaptation, poor SaaS application response time is no laughing matter. It costs businesses millions of dollars in lost productivity and starves providers of revenue, to say nothing of the dramatic toll that continues to be taken on user sanity.
There is plentiful research out there on how long application response times result in bad consequences, and we’ll get into some of the science shortly, but this is an area where you don’t need to look beyond your own experiences to validate the problem.
When you’re shopping online and you get a spinning wheel while the site pulls up the results of your search, how long do you stare at it before you’ve opened another tab looking at a competitor’s site? How long do you wait for that online document you need to edit to load before you’ve context switched and gone to check your inbox? If you’re like most of us, it’s a matter of just a few seconds.
In fact, studies consistently show that after just a second or two of delay, a user’s attention wanders from the task at hand. This type of finding would be a perfect opportunity to bemoan what our modern instant-gratification, thousand-choice, social-media-driven environment has done to our collective attention spans. However, we see consistent findings going all the way back to a landmark IBM study in 1968.
It’s Only Natural
It seems that this need for feedback delivered on a reliable and comfortable cadence is simply part of who we are as humans. There is a natural flow to a human conversation, and the length and location of pauses is as integral to that flow as the downbeat is to a musical performance. In a normal conversation, pauses typically last between one-quarter and one-half of a second. A four second pause in a conversation has been shown to significantly decrease listener comprehension, in addition to feeling excruciatingly long.
A comfortable rhythm in a human conversation also resonates with the participants in a way that helps to form a positive social bond. On the other hand, a disruption in that rhythm can be alienating, and leave both parties with a sense that they just “didn’t click”.
That’s just the way we are wired, and those connections are still firing when our conversation is with an application rather than a person. Long wait times for SaaS application response, unreliable pauses, and a general disruption in our conversational workflow will leave us unfocused and irritated. After all, we’re only human.
Changes in application implementation technology have no impact on this primal cadence. In 1968, the IBM study identified a loss of user concentration if the delay between feeding in a punch card and getting a response that the data had been received exceeded two seconds. In 2017, a Google study found that most mobile users abandon a website that takes longer than three seconds to load.
Jakob Nielsen is a titan of usability consulting. When asked if the response time expectations were different for web-based applications, he responded that “[t]he response time guidelines for web-based applications are the same as for all other applications. These guidelines have been the same for 46 years now, so they are also not likely to change with whatever implementation technology comes next.”
No Free Pass For SaaS
SaaS providers should take note that there is no free pass given by users simply because their application is being consuming through the Internet. “Oh, it’s ok that my important project updates were lost! After all, this application is in the cloud”, said no user ever. Until human nature changes, human expectations won’t change either, and there are reasonable human expectations around SaaS application response times.
Part of the calculus of profitable SaaS operations is the ratio of Customer Lifetime Value (LTV) to Customer Acquisition Cost (CAC). As a SaaS provider, you want that ratio to be as high as possible. There are many ways to maximize it, and one of the most fundamental is simply to hold on to a customer as long as possible. The Harvard Business Review notes that the cost of acquiring a new customer is anywhere from 5x to 25x that of retaining an existing customer.
A customer who is dissatisfied with the service levels delivered by a SaaS application is at a high risk of attrition, also known as customer churn (for a more in depth look at churn see this blog post). Engaged, productive users are happy users. Happy customers are made from happy users. And happy customers equate in a very direct manner to the profitability of the provider of the service. SaaS application response time is a major factor in creating those happy users.
None of this will come as any surprise to a successful e-commerce business. They have long since recognized the value of lightning-fast responses to their customers, and that need has powered the massive and continuing growth of Content Delivery Networks (CDNs). CDNs do a phenomenal job of delivering static content, like an inventory page for an online retailer, very quickly and efficiently. They do this largely by caching the website’s content on their own servers, which are geographically distributed and therefore both logically and physically closer to their customers.
The problem with this approach for SaaS providers is that their content isn’t static. SaaS application response and interaction is inherently dynamic and bidirectional, and that means that caching simply isn’t an option. The customary solution has been to shard their application and deploy it into a number of regional PoPs. The closer you get to the users, the better the performance. This comes with a huge price tag, though, both in up front and operational costs.
Internet Overlays And SaaS Application Response Time
There is another way that providers can optimize their global SaaS application response time, and that is through the deployment of an Internet Overlay Network.
Internet Overlay Networks like Teridion enhance Internet performance through dynamic route optimization and protocol acceleration. They accelerate all traffic, including bidirectional and dynamic traffic, to any user anywhere in the world, as a turnkey service. Deployment is easy, and there’s no upfront cost.
If you’re a SaaS provider looking to keep your customers engaged, happy, and loyal, it’s worth taking a deeper look at Internet Overlay Networks. This short video from Teridion is a great overview of the SaaS performance problems presented by the Internet, and what can be done to solve them. If you want to see how Teridion can radically improve SaaS application response time, it’s easy to start a free trial.