/ Getting Started

Personalizing the user onboarding for your SaaS with Segment, Clearbit and Artificial Intelligence

There are numerous issues that can lead customers to leave a business but studies have found that nearly nine out of ten customers leave due to poor experience such as a poor onboarding process. As we are currently living in an era of customers demanding exceptional experiences, imagine if your business could predict the next-best action for the user, at each point in the user journey, in order to maximize the retention rate. Well, you can.

Traitly is a company powered by Artificial Intelligence (AI) meaning that they have developed a platform called Traitly Predict that predicts how to nurture customers through personalized step-by-step onboarding for each user. The platform encourages each user to perform a suggested next step which strengthens the retention and onboarding rate. By predicting the next best action for each user, Traitly enhances the experience of the users whilst you can lower costs associated with attrition and instead benefit from growth.

In order to make such predictions Traitly acquires data from Segment and Clearbit, two platforms that gather and enrich your raw data before sending it to Traitly. This article aims to outline, step-by-step, how you can optimize the user onboarding journey for your freemium-to-premium SaaS business by leveraging data and AI.

How you can make use of data that lives inside Segment and Clearbit

You cannot understand very much about your users unless you track who they are and how they behave. This data also needs to be in a standardized format, otherwise you will end up spending hours and potential days trying to clean up your data before you can even begin analysing it. Luckily, there are a bunch of tools that can help you with this.

Leading SaaS companies use Segment, which is a user data platform that collect data from you and your users. This means that Segment collect your data from each platform you are using and in turn send the data to hundreds of tools such as Traitly, for analytics, marketing and data warehousing to assist in driving growth and revenue. This data enables Traitly to gain a complete picture of how your users actually behave.

Furthermore, there are so called specifications within Segment that are important such as projecting meaning to the data through APIs. APIs are a set of requirements of how one application interacts and communicates with another. The APIs that Segment uses include Identify, Track, Page, Screen, Group and Alias. Each represent different information about a user. Although, the most vital APIs that provides Traitly with valuable data are:

Identify: Who is the user?
This is the first action to be made when a user has registered, just logged in or updates information such as changing address. It is important as it ties a user to their actions and records their traits.
{ "type": "identify", "traits":{ "name": "Peter Gibbons" "email": "peter@initech.com" "plan": "premium" "logins": 5 }, "userId": "97980cfea0067" }

Track: What is the user doing?
It refers to how you can record each user’s actions as well as any properties describing the action. These are defined as events and properties are extra pieces of information you can tie to such events.
{ "type": "track", "event": "Registered", "properties": { "plan": "Pro Annual", "accountType": "Facebook" }

Group: What account or organisation is the user part of?
A user can be more than a single person and be a part of a group such as a company, organisation, project or team.
{ "type": "group", "groupId": "0e8c78ea9d97a7b8185e8632", "traits": { "name": "Initech", "industry": "Technology", "employees": 329, "plan": "enterprise", "total billed": 830 }

Once Segment has gathered data, many companies use Redshift, Amazon’s data warehouse in the cloud, to store historical user data - that is, how users interacted with your app in the past. Traitly can connect directly to your Redshift database and access past events such as users' behaviour. Segment Webhooks by contrast, is used to send a real-time data feed of everything your users do inside your app to Traitly, for example if a new user register on your website, Segment will automatically notify Traitly. The large quantities of data from these platforms contribute to a good view of your company’s past, present and future.

Once Traitly has obtained a well-rounded view of your company and your users, Clearbit can provide additional insights. It does so by accessing up-to-date relevant background information such as each user’s location, employment titles, number of employees (where your user is an organisation - B2B), employee names, industry, and so on. Such information is valuable for Traitly if, for example, the goal is to predict the users’ behaviour in a specific region, city or those holding various job titles.

Let’s take a concrete example which shows how Traitly can utilize raw data augmented by Clearbit:

Suppose you operate a customer support SaaS company. It is your job to provide a great platform and help your new users get up and running successfully. Typically, you will have two types of users:

  • Customer service agents
  • Customer service managers

Each of these users will have a separate role. The kind of actions each user role needs to undertake are likely to be different. For example, the customer service manager will need to invite the customer service agents team, set up agent performance tracking, and make a payment when the time comes.

Customer service agents, on the other hand, are on the frontline. It is their job to figure out which messages need a reply, how to prioritize incoming queries, reply to customers, create an email signature, and tag messages.

While both users roles can feature in a single account of your customer service SaaS app, it is clear that the optimal user journey looks different for each role. Agents do things managers do not and vice versa. With this in mind, you can create personalized user journeys. By instructing Traitly to optimize for each user role separately, Traitly can analyse and learn the optimal sequence of events for each user. Traitly’s AI recommendation platform can guide the user through your app in real-time, always recommending the next step bringing them closer to becoming a paying user.

So, how does Traitly use the data acquired from Segment and Clearbit?

Now that you have some understanding of what Segment and Clearbit can do, a clarification of how Traitly uses these platforms in practice will be presented. Here are the six steps:

1. Create a Goal
The first step when you decide to benefit from Traitly’s AI, is to decide what you wish to optimize for your users - we call it a goal. A goal can be, for example, to get your users to convert from freemium-to-premium within the first 30 days of their free trial. You can also create fine-grained goals, for example, optimize the user journey for those users who have a specific role in your software - support agents vs. a support manager being one.

2. Connect Data
Once you have decided on what to optimize, Traitly receives and uses your Segment real-time and historical data enriched by Clearbit. Typically, it takes just a few minutes to successfully connect your Segment data. To make this process more clear, an example is illustrated.

Segment sends a message to Traitly whenever a user does something inside your app. This is facilitated by Segment Webhooks. You need not concern yourself too much about this, beyond knowing that you have to copy and paste a Web address from Traitly to Segment (for more information on how to Connect Your Segment Data). Traitly automatically handles the rest of the work for you.

3. Data Processing (or “data formatting”)
This step refers to converting the data into usable and desirable forms. Here, Traitly understands your users’ past and current behaviour taking the time frame and events into consideration. This makes it possible for Traitly to interpret and provide insights into the data and fully comprehend how your users act.

4. Receive recommendations for your users
When the data has been processed, the desired forms of data is then further calculated and predicted through Traitly’s AI algorithm. The algorithm generates next-action recommendations for each of your users in real-time, much like the jobs to be done framework. This means that Traitly personalizes the user journey based on what each user has done so far and what Traitly predicts they need to do next. Personalization is automated with a specific goal in mind, for example, your desire to get users to convert to your paid pricing plan within the first 30 days of their trial.

To illustrate an example of such actions, imagine an app that offers a music service with a 30-days free trial. During this period, the user will receive various personalized recommendations. This means that if Traitly’s algorithm identifies that a specific user is not very likely to convert within the free trial period, then the recommendations for this user will include simple tasks, such as ‘Search for a song’. Once this step is accomplished, it will not be shown again instead a slightly more extensive recommendation will appear, such as ‘Create a playlist’.

If a user on the other hand is very likely to convert to the premium version, then more comprehensive recommendations might be offered, such as ‘Invite a friend’. This is just a simple example - the AI automatically understand what the best recommendation is for each user. Such step-by-step personalised recommendations increases chances of conversion from freemium-to-premium.

6. Retarget Users
The last step is retargeting users through platforms such as Facebook, Twitter and Hubspot. It refers to effectively engaging with users who have signed up for your app but have not yet converted. The aim is to engage users in an effort to improve their odds of converting.

Facebook is a great channel for retargeting due to its multiple users and Facebook dynamic retargeting ads enables to target users with relevant ads at all stages of the sales process. Through these ads, it is important to give the user a reason to return and finish what they started. To illustrate an example, think about the music service again.

If a music streaming service offers a 30-days free trial period, users will receive Traitly’s AI-tailored recommendations on the right hand side of the Facebook page, constantly reminding the user to complete the next step such as ‘Create a playlist’. Traitly will detect if the user completes the recommendation and will calculate a new recommendation based on the user’s behaviour. This method can similarly be used on Twitter as well.

Hubspot is a platform in which companies can create mostly email marketing messages. By connecting Traitly with your Hubspot account enables you to realize which user to send your marketing messages to and in what order. This imply that once your email marketing messages are created on Hubspot, Traitly sends dynamic lists of users that are to be connected to the messages created so that the right user receives the right message in the right order. The message aims to informs the user what specific tasks that needs to be completed next. For example, the customer service manager will receive a message that informs how to invite the customer service agents or reminding the manager to set up agent performance tracking, which ever is predicted to be the next step.

Why you need Traitly

The best way to avoid user attrition is to truly know the users. Because if you know your users and how they behave, the user onboarding process can be personalized increasing chances of conversion and builds loyalty. Furthermore, retention is crucial as it is much less expensive to retain existing users than it is to acquire new users. SaaS companies, like yours, spend seven times more on user acquisition than user retention, making it clear that user attrition is costly for businesses. Thus, the benefit of predicting how to retain users and get them onboard for the long run can decrease costs which in turn increases the chances of growing your business.

If you would like to try Traitly Predict today please request a demo and the Traitly team will reach out to you.

Personalizing the user onboarding for your SaaS with Segment, Clearbit and Artificial Intelligence
Share this

Subscribe to User Onboarding - Conversion Optimization | Traitly