Need to integrate a few apps to solve a business problem? There is recipe for that
- 30 June, 2016 05:08
Vijay Tella has been neck deep in application integration technology for years, first as the SVP of engineering at TIBCO, the company that introduced the information bus, and then at Oracle, where he helped launch the company’s booming middleware platform. Today Tella is founder and CEO of Workato, a company that is putting integration tools directly into the hands of app users. Network World Editor in Chief John Dix recently caught up with Tella to learn more about how he is trying to democratize the world of app integration.
Does the company name mean something?
It stands for work automation. We’re all about making it easy for business people to integrate their apps. There are so many amazing apps in every category for every type of business, from small businesses like restaurants to Fortune 500 companies. Often these apps are picked up not by IT people, but by business users. They’re relatively inexpensive and each one does a great job for the specific thing they do.
As a group, however, they become problematic because your customer information, your product information, order information, etc., all of this gets fragmented across different applications. The coordinated information you need to address the important things going on in your business becomes challenging when your work is spread across so many apps.
Individually, the apps are productive, but as a group there are serious productivity issues with apps. That’s the problem Workato has set out to solve.
What types of apps are you talking about?
Everything from Salesforce to Netsuite and QuickBooks and, in bigger companies, ServiceNow and SAP to Oracle. It’s every possible kind of business app. We work with pretty much any app that has an API exposed, but we also support those that support comma separated values and those that support REST (representational state transfer) style interfaces. We connect with thousands of apps.
I was involved in creating some of the first technologies in application integration. I was part of the founding team of Tibco where we created something called an Information Bus that helped disparate apps talk to each other. It became a gold standard for application integration a long time ago. And I helped start the Oracle Fusion Middleware platform, a multi-billion dollar business today. The reason we have done Workato is to bring the power of Tibco and Oracle to users of all of these cloud apps, but enable them to do the integration themselves with little effort.
Give us an example.
Take a small business like a restaurant using an iPad at its point of sale to capture sales information and an accounting app like QuickBooks for their books. The sales information from the point of sale needs to be reconciled and saved and summarized in the general ledger and the data structures behind this daily sales information and general ledger are complex and you need to be able to map between them.
Not only that, when you are syncing between these things, the sync needs to be robust. If you end up ringing up the sale twice or you had duplicate data or you have some incorrect or bad data that you propagate from your point of sale to your accounting, it messes up your system in a really bad way. The level of integration required needs to be solid, it needs to be robust, it needs to be able to handle the complex logic. And often the people that are using it are nontechnical. They don’t even want to know about this stuff, they just want it to work.
As you get into larger business and more complex apps like Salesforce, NetSuite or SAP, the requirements for enterprise grade, robust integrations escalate.
The problem we set out to solve is to bring the power of robust enterprise integration to nontechnical business users, from small businesses to enterprises. In small businesses, it is mostly about automating manual work. In larger companies it is about automating cross functional workflows across different departments.
Integrating everything under the sun seems impossible on the face of it, so describe how you work that magic.
There are literally thousands of apps out there and millions of work groups using those apps. One of the key things we did was embrace a community dynamic.
You’re familiar with GitHub? It’s an amazing community for developers, where if I want a charting tool for a program, on GitHub there are probably 20 public domain charting tools I can consider. So I pick one, copy it, make any variations I want, and that’s it. You can have private code on GitHub but a lot of the innovation happens in the public domain there. GitHub is a multimillion dollar company, but the value of GitHub is not what the GitHub engineers have done; it’s the community that makes it an amazing thing.
The approach we’ve taken is very much like GitHub. The key concept at Workato is a thing called a recipe. A recipe is an English-like set of instructions. It’s a conversational way of describing how apps should work together, what information from what app is going to what app and in what condition and how it is transformed, all that stuff. These recipes are designed to be fundamentally shareable and, by default, recipes are public on Workato. Earlier this month we passed 100,000 public recipes.
Each recipe is like a little mini-app for integrating specific apps. If you’re looking to integrate, for example, QuickBooks or Netsuite with Salesforce, we have hundreds of recipes that do that. They are variations of each other. You consider your process and then go find something that works for you exactly or something that is close enough that you can tweak to your liking, because it is English-like and very conversational.
It is designed to be easy to create new recipes, but it’s even easier to tweak an existing recipe. The goal for our community is to have 1% of people creating original recipes, 9% making variations of existing apps, and 90% simply finding what they want and running with it. Today, about 70% percent of people who come in find an existing recipe and go with it and 30% are either creating new or making changes to existing recipes.
Do you create any of these recipes yourself?
Partnerships have always been a big part of Workato. When you’re capturing orders from Salesforce and generating an invoice in Intacct, it’s a specific scenario that people at Salesforce and Intacct know. We partner with several dozen ISVs, including Intacct, ServiceNow, Zendesk, Marketo and QuickBase, all sorts of companies, and we work with them to create a set of curated recipes that are good seeds for the community.
So we create some and our ISV partners create others. Some of the most active people at Workato are these consulting partners, big enterprises to small consulting companies that create recipes. They can keep the recipes private or open them up for anybody to consume. The benefit of opening them up is when other people improve those recipes you also get the benefit.
When you look at recipes at Workato you can see how many companies are using a particular recipe. When we see some core scenario emerging, we feature some of them. So if you are searching, say, for an app for organizing a sales event and you want to take all the attendant information and load that into a marketing automation tool like a MailChimp, we’ll show a prioritized list of recipes.
We have a lot of private recipes in the system too. When you’re creating a recipe you feel is proprietary to how you’re doing business and you don’t want to share it, you can keep it private. But when you have your recipe in the public domain it gets more use, and the more it gets used the more robust it becomes and the more it benefits you.
How about giving us a few more common use cases.
Take marketing automation. American Kennel Club uses Marketo to organize campaigns for different dog products and they have an ecommerce engine called Shopify. They use Workato to reconcile who is buying products with the campaign used. That means their follow up is smart. They’re not going to try to sell something
to someone that has already bought the product. It’s the basic closing of the marketing loop.
But we have apps that are popular for everything from employee onboarding to accounting, finance, marketing, sales and field service. If you can think of an app scenario that needs integration, we can probably address it.
Another scenario that is emerging is use of chat products like Slack and Hipchat. Workato created a chatbot called Workbot that lives inside products like Slack and collects information for you. So if you’re using Slack for support, for example, when a customer trouble ticket comes in and you need more information about the problem or the customer or both, and that content is in other apps like Salesforce, instead of users having to hop across multiple apps this Workbot goes and gets the context. You can also get charts and graphs from these apps in Slack and, more importantly, take action right from Slack. We see this as one of the fastest growing use cases for Workato.
There are some Platform-as-a-Service tools emerging that you can use to do this kind of integration work. How do you stack up against something like that?
Bigger companies like Salesforce and ServiceNow have their own platforms. They’re not just apps. They have a developer platform that you can use to extend Salesforce. But there are two types of people in companies: people that can code and people that can’t. For people that can code, they can use developer tools because the apps have APIs.
We abstract all those APIs and take the technology out of the integration, but if you’re a hardcore developer, you can directly code integration. But there are a few issues with that. One is when you code these integrations they are fixed, meaning you’re hard coding a scenario. When you’re scripting it on Workato you can tweak and change a line of logic to make the recipes a little spicier or whatever. You can just tweak things a bit.
And another thing is the people that are typically buying these apps are not developers. They are business people. They are the head of marketing or HR or admins for these apps. These PaaS platforms are not relevant for the 90% audience we’re talking about.
Are you strictly targeting shops that don’t have coding resources?
I wouldn’t say strictly. We have a lot of consulting companies who have developers in-house that use Workato because of the productivity aspect. We have some power features that are very popular for developers because they can do the same thing ten times faster.
So who do you compete with?
There are two types of players. People that have solved this problem deeply and thoroughly at a level of functionality that makes it secure, eliminates duplicates and is robust. These are companies like Tibco, IBM, Oracle, MuleSoft, Informatica, etc. Their products tend to be expensive and you need to get services from companies like Accenture to use them.
Then there’s another set of players that come from a consumer background, “If this then that” (IFTTT)-like tools that make it easy for consumers to do things like, say, link an Instagram feed to a Dropbox folder. Gartner and Forrester call those citizen integrator tools. We are the one product that spans the spectrum of citizen and enterprise integration.
But the biggest competition is people doing stuff manually or simply suffering with the consequences of not doing it.
How do you charge for it? Is it a subscription?
It is a subscription based model, and the subscriptions are primarily based on the type of apps you use. We have targeted integrations that start at $49 per month, but if you’re using small business apps you pay $99 a month and you can connect any other apps, you can run any number of transactions or integrations through them. We don’t limit you.
For midmarket apps like Marketo, a marketing automation tool, the logic tends to be more complex so support requirements tend to be higher, so we have a Pro plan which is $500 a month. And then we have Enterprise. If you are using Marketo Enterprise Edition or Salesforce Unlimited or using on-prem apps like Oracle or SAP, those start at $2,000 a month.
This cost is an order of magnitude lower than the MuleSofts and the IBMs, the traditional big companies. When you have highly complex, high touch products, they have to be expensive.
Our business model is based on high volume and the cost is based on the type of company you are and the type of apps you use. We also have special pricing for nonprofits.
If I’m a customer and I find a recipe that’s going to help me, do I download code or am I just spinning something up in the cloud and having you return what I need to that other program?
What you do is you go to Workato.com and say – “Hey, I’m looking for Eventbrite and MailChimp since I’m doing an event and I need to connect those two.” There’s a search bar. You type those and we’ll show you the list of recipes for Eventbrite and MailChimp. There’s a summary that we use that have little cards.
With the card-level view you can look at 10 cards at a time and, when something looks right, you click on that, get a detailed description of exactly what it does, and if that’s what you want you click “Get This Recipe,” which copies that into your account. Once you get the recipe you spin it up to connect to your Eventbrite account and your MailChimp account and then we have a test button so you can look through a couple of loads of data from MailChimp to Eventbrite to see if it looks right. If it does, you simply hit start and it’s done. From that point until you stop it, it will keep syncing all the time.
We have gone through extensive third-party security and privacy certifications around this. We created this to be as secure as the enterprise apps integrations we’ve done before.
How many customers do you have?
We passed 13,000 in June.
How do you judge your success? Is it the number of processes you’re handling in a given month, the number of customers, revenue, or a combination of all those?
There’s this thing we call FTA, first time activation. How many people have achieved activation and are running recipes? That’s the number one metric I look for. We have these dashboards and that’s the one I look at every day and I can click through and see how they’re doing.
The second metric is really the number of transactions. A single recipe can have on average about 10 steps and each step takes between 2-3 minutes of people’s time. For me, the number of transactions that we are doing is directly representative of how much time we are saving for people. And third is the revenue.
Workato is my fourth startup. One of the reasons for doing Workato is this is a really big problem and can make a real difference in taking some of the worst parts of people’s jobs out of their hands.