SurveyMonkey CEO finds answers for the enterprise
- Summary:
- SurveyMonkey, the long-established online survey tool, has been on a five-year journey to rearchitect for the enterprise market. CEO David Goldberg tells me why
One question was top-of-mind: surely collecting answers to questions is something that's really easy to build into an application. Why would there be an enterprise market for a separate cloud service that's solely dedicated to collecting surveys? Here's how Goldberg responded:
If everyone's going to ask the same question in your application and they're going to get the same data output, sure you probably don't need us. But that's not the way the world works.
What people find is that it's actually a lot harder than it looks — as soon as you get to more than one question, or you have a lot of responses and you want to be able to do something with those responses, all of a sudden the complexity gets a lot harder.
People don't want to ask one simple little question, they want to ask complex questions — and they want to have that data be able to be manipulated and analyzed. So you've got to create a structure around that. That's where there's a lot of hard work.
Another advantage of SurveyMonkey is that anyone can use it to build ad-hoc surveys or analyze results without needing any IT expertise.
People are using the logic to run [respondents] through a series of if-then statements and come out at the right place. It's all about capturing that data. It is a form of programming — but we don't want to tell people that, it might scare people away!
Adding 'why?' to 'what?'
Goldberg argued that surveys are important because, although it's possible to collect huge amounts of contextual data these days, you can't reliably infer what's going on in the minds of your customers or users without asking direct questions. A survey adds a subjective dimension to the objective facts of big data, he explained.
People talk about big data all the time. They often mean, 'implicit data collected while people are doing something else.' [At SurveyMonkey] we use that type of data to run our business. We look at all the stuff people are doing in our application and understand where they're getting stuck or where the upgrade triggers need to be moved to. There's huge value to that.
But we also think big data is about 'why?'. You get a 'what?' from all that implicit data, but sometimes you need to ask people explicitly, 'Why did you do this?' or, 'What do you need from us?'. That's where the asking questions part comes in.
I think you'll see lots of examples from us where we're marrying the implicit data with the explicit questions. We're not going to be collecting, necessarily, all that implicit data — there's lots of people doing a great job of that. You want to marry our data with that.
The integration to Salesforce is an example of an application that provides contextual data, to which SurveyMonkey adds direct feedback. This has two advantages, said Goldberg:
First of all I don't have to ask the customer a lot of questions that are extraneous that I already know the answer to — it's in their customer record. So I can keep the survey short, which is really important in a world of people having limited time and attention, and taking [surveys] on mobile devices. The shorter the survey is, the better your response rate.
The second advantage is, inside of Salesforce, being able to sort and analyze that survey data in the context of all the other data. Maybe I want to see it by support rep, or by day, or by type of customer.
Eliminating Excel
This type of analysis is the main reason integration to a survey app has been one of the most frequently requested features in Salesforce, said Goldberg:
That's why this is something people really want, because they're doing it already. They're running the customer survey in SurveyMonkey and they're using Salesforce, but they were having to marry it all together on the back end, in Excel.
That need to eliminate Excel as a means of joining data from multiple sources is what's motivating a lot of cloud app adoption, he added:
People talk about consumerization of enterprise, SaaS — a lot of it's just, how do we get rid of Excel? It is a multi-purpose wonder tool but it requires a lot of work and a lot of manipulation, it's subject to all sorts of problems and it's hard to automate.
Integrating to Salesforce is just the first of many others, said Goldberg.
At the end of the day, what people want to do, they want to collect data from customers, from employees, from partners, from vendors, whatever that source is, and they want to understand that data contextually with whatever other applications they use to manage those relationships.
Longer term we have a goal for people to build on top of us as well ... we want to have a pretty open set of APIs, platforms, where people could build.
Enterprise journey
It has been a long journey to get to the point of being able to start putting this enterprise platform strategy into place. SurveyMonkey was founded in 1999 by Ryan Finley as a side project while a college junior. Cloud-based before people called it cloud, freemium before the word was coined, and at the forefront of consumerization of the enterprise — the company was a pioneer in many ways.
When Goldberg joined as CEO in 2009, the company was the undisputed leader in its category and highly profitable, but had just 12 employees and was still architected for individual use. Today it has 450 people and is valued at more than $1 billion after a refinancing that allowed some original investors and employees to withdraw equity. Goldberg recounts the journey since then:
After I hired a team the first thing I had to do, we had to rebuild our whole product and technology architecture. We inherited a product that scaled and worked but was not very flexible. We're almost done with that, but the fruits of that are now coming to bear.
We had to get our technology architecture to the point that we could really do this the right way. We had to have an enterprise version of the product which has all the authentication and security — you've just got to have that stuff. It's not something you can just bolt on.
Mobile use cases
Now that the API layer is in place and the Salesforce integration has been done, a mobile developer kit is being prepared for launch. An obvious use case is to collect feedback from mobile app users, says Goldberg:
Right now, there's not a good solution for in-app feedback from customers that's easy and cost-effective. You don't really want to send people to an app store to tell you that they have a problem with the product.
Another use case has been developed with mobile phone maker Samsung, in which a security guard uses voice input to report in while doing their rounds, while the phone records their location. It's another angle on the notion of combining contextual data with survey responses. The voice-driven survey app becomes the digital equivalent of ticking boxes on a control sheet carried on a clipboard. Goldberg comments:
We're constantly amazed by what our customers actually come up with. We're giving them a platform to collect data that they make decisions from. What those decisions are, what's that data they want to collect, it's all over the place. Anything you can think of, people have decided to use us for. As we start tying these things together, it'll get even broader.
We look at it as the collection of data. Any mode that our customers want to interact with their [customers] or their employees, we've got to be in those places. That could be inside of other applications, it could be through email, on a web page, in mobile, whatever mode somebody's in, we need to give people the ability to collect that data.
My take
SurveyMonkey's journey is a reminder of the many differences between a standalone, single-user application and the same functionality delivered in an enterprise context. The enterprise environment brings requirements for user administration and compliance as well as an API layer that allows for interactions with many other applications.
Now that the provider has completed its rearchitecture, it has to justify its existence by demonstrating that there is a market need for survey functionality as a separate, plug-in service.
In the old world of enterprise IT, the tendency has been to harness as much functionality as possible in a single platform — providing 'one throat to choke' for support, licensing and so on. SurveyMonkey's ability to thrive as a separate service will show us how strongly such concerns are still an issue in the cloud.
Disclosure: Salesforce.com is a diginomica premier partner.
Image credit: © Marek - Fotolia.com