DEV Community: Zack Khan

Announcing Hightouch Audiences: Enabling Marketers to Self-Serve their Data

Zack Khan — Tue, 31 Aug 2021 18:23:31 +0000

You have this amazing idea for a marketing campaign: sending your customers this congratulatory Gif after they complete their first purchase. Genius, right?

You have the copy written. All that's left to do is to get a CSV with
all of your customers that made a recent purchase: sounds simple. You
ask your lovely data team for the CSV, and bam! You're told to add a
JIRA ticket in a never ending backlog of 100 other data requests from
across the company. The earliest they can get to your ticket is next
month. What do you do now?

We've all been there before. But why is this the norm? Here at
Hightouch, we believe that everyone should be able to use data to
personalize the experience of their customers. Today, we move one step
closer to that vision with our latest product for Marketers: Hightouch
Audiences.

Goodbye CSVs, Hello Hightouch Audiences

Hightouch Audiences allows marketers to define audiences to target for
marketing campaigns and sync those audiences to any of their marketing
tools like email or ad tools: no engineering favors required. As a
marketer, you can use our Visual Audience Builder to visually define and
filter audiences in an intuitive UI, without needing to know SQL.

Once you define an audience, it will continually sync new members of
that audience to your tools: no more having to upload a new CSV every
week.

Use Cases

Previously, the data warehouse was only accessible by SQL-savvy team
members. With Hightouch Audiences, the power of the data warehouse is democratized so that anyone can personalize customer experiences for use
cases such as:

Lifecycle marketing: Send lifecycle marketing campaigns to customers across any channel as soon as they invite a friend or abandon a shopping cart
Target Your Paid Ads: Increase your ROAS by retargeting customers who visited your pricing page or excluding customers who already purchased
Create lookalike audiences: Find new customers similar to your existing high value customers in all of your ad networks like Facebook and Google
Send Conversion Events: Send conversion events to any ad network to optimize your targeting, reduce your CAC and provide enhanced ROAS reporting

So, how does it work exactly?

You can think of Hightouch Audiences as the glue between marketing and
data teams.

First, data teams define models and relationships. With just a SQL query or dbt model, data teams can define the data that Marketing can access (columns, tables, etc). Hightouch is flexible and supports any data available in your warehouse: Users, Accounts, Workspaces, Products, etc.
Next, marketing teams create audiences on top of data models. Marketing teams can visually filter audiences based on any properties or events. For example, you can choose an audience of users who purchased a certain item or live in a certain city.
Then, audiences are synced to marketing tools. Audiences can be synced to CRMs, ad networks like Facebook, marketing automation tools like Marketo, lifecycle marketing tools like Braze or Iterable, email tools like Mailchimp and 60+ other tools.
Finally, marketing teams can run campaigns on those audiences. You can deliver personalized messaging to those audiences in your marketing tool and channel of choice, such as email, SMS or in-app.

Where does Hightouch Audiences sit within my marketing stack?

Hightouch Audiences acts as the central data source of your marketing
stack that enables you to run multi-channel campaigns in all of your
tools. For example, you can define an audience of users who visited your
pricing page. Then, you can sync that audience to your email tool (ex:
Mailchimp) for a personalized pricing email and retarget them on ad
networks (ex: Facebook) to get them to convert. All you need to do is
define your audience once, and Hightouch takes care of making that
audience available in 60+ different tools.

What makes Hightouch Audiences unique? How does it relate to a
CDP?

Most CDPs (Customer Data Platform) and audience builders like Segment
trap you in a rigid data model, lock you to use just Events and Users,
store your data, and hit rate limit issues at scale. The key difference
between Hightouch Audiences and any other audience builder is that
Hightouch is more flexible. How? Because it works on top of your data
warehouse, which is the single source of truth for your business. Your
warehouse has the full picture of your customers, including product
actions, billing information and even 3rd party data sources.

All this context on your customers allows you to create powerful
audiences. You can even use Custom Objects specific to your business
(like Workspaces, Accounts, Products, etc): for example, filtering all
customers who bought a specific Product or belong to a high value
Account. And your data stays in your warehouse, keeping your precious
customer data safe within your systems (which is especially needed for
regulated industries like Fintech and Healthcare). That's something a
CDP (Customer Data Platform) can't do.

However Hightouch does not help with event collection: you can still use
a CDP or solutions like
Snowplow for that.

Is Hightouch going to replace BI tools like Looker, Amplitude or
Mixpanel?

BI tools like Looker have also made data accessible to more people in a
company, but at the end of the day, they are focused more on analytics
(hence, "business intelligence") than activation ("putting data to
work"). In order to activate data in BI Tools, it's still fundamentally
all about CSVs. Hightouch Audiences is even more accessible given our
intuitive Visual Audience Builder, and even more powerful and automated
than CSVs because it connects directly to your marketing tools. This
saves you time from doing manual work and helps you automate your
campaigns.

What data does Hightouch Audiences use? What setup is required?

Hightouch Audiences requires that you have a data warehouse or data lake
that has event data on how your users are interacting with your product.
Hightouch does not replace your existing event tracking or ELT
workflows: in fact, it enhances it. The
more data you make available in your warehouse (such as billing data
from Stripe, sales data from Salesforce, etc), the more powerful your
audiences can be.

Does Hightouch support real time?

You need to reach customers when your message is most relevant to them,
like the example above of sending a gif as soon as a customer
completes a purchase (not 10 minutes later). Hightouch supports
real-time use cases: it can send data to marketing tools as soon as an
event lands in your warehouse, or you can use a streaming data source
like Segment, Kinesis or Kafka.

How do I try out Hightouch Audiences?

Learn more about Audiences and book a demo here.

What is Operational Analytics?

Zack Khan — Wed, 25 Aug 2021 02:15:38 +0000

Operational Analytics

It's common to hear teams talk about the importance of "data-driven decision-making". Once a lofty aspiration, innovations in data warehouses and BI tools have made it simpler and cheaper than ever to actually make sense of data. But there's an unsolved challenge - insights gathered from analytics are only valuable once they're actually used to make a change in the business that moves the needle. This is sometimes referred to the "last mile of analytics."

Without that elusive last mile, analytics is at best a reactive report card for your business, and at worst, a waste of time. At Hightouch, we've worked with hundreds of companies who struggled with the last-mile of analytics problem: all of their important data lives in the warehouse, reporting is solid, but it's too hard to take action on that data.

Operational analytics is an approach to analytics that shifts the focus from simply understanding data to actually putting that data to work in the tools that run your business. Instead of just using dashboard data to make decisions, operational analytics is about turning insights into action - automatically.

The two uses for data

Every company generally uses data in 2 ways:

Operations: using data to actually "do things". For example, triggering an email when a new customer signs up or makes a purchase.
Analytics: using data to understand what's going on in the business. For example, building an executive dashboard with KPIs across sales, marketing, finance, etc.

Operational data is all about syncing data between systems to do things like communicate with users, bill customers, alert employees, etc. Analytics is often see as one of many "destinations" for the operational data pipeline.

The persistent challenge with operational data is that it's not easy to get your various tools to "talk to" one another. For each pair of tools, you need to figure out how to get data to flow dependably and accurately between them. If you've ever gotten an email addressed to first_name, you've seen this notorious challenge rear its head.

Some other examples of operational data workflows:

A B2B software company syncing product usage data to a CRM so a sales rep knows when to reach out to a customer.
An ecommerce company syncing purchase data to an ad network so that recent purchasers don't get targeted for something they already bought.

Analytics, on the other hand, is about bringing all kinds of different data together and visualizing it in a way that paints a picture of what's going on in the business.

The beauty of analytics data (which turns out to be the key that unlocks operational analytics) is that it's often the only realm where different datasets live together harmoniously - most often in a data warehouse. Analytics data is tied together neatly through models that form the foundation of the digestible, contextual charts that analytics tools provide.

Thanks to innovations in data warehouses and the surrounding ecosystem, bringing data together for analytics has never been easier or more cost-effective.

The analytics layer is a hub, not a spoke

It just so happens that what was previously thought of as the "analytics layer" turns out to be the perfect foundation for operational data workflows, and an antidote to those challenges associated with getting systems to "talk to" one another.

As opposed to creating point-to-point connections between tools ("spokes"), companies are now beginning to use the warehouse as not just the foundation for their analytics, but as the "hub" for all operational data workflows. This is operational analytics.

There are a few reasons why the "analytics layer" is the ideal hub for operational workflows:

It's simple to aggregate and integrate data in data warehouses; it's what they're built for. Teams can easily bring customer data, billing data, employee data, and other datasets together into the fabled single view of the customer, a promise made by many SaaS vendors who haven't really delivered. Once the data's in the warehouse, the path of least resistance is to just send that data out where it needs to go.
Security: companies own their warehouse, so data never has to leave your purview and fall into the hands of yet another vendor.
This approach breaks down silos between data teams and business teams by creating a clear handoff: data teams own the raw data and model it into clean data, which empowers business teams to own the management and sync of that data to the tools they need to run the business.

This approach is dramatically changing how companies think about analytics.

Analytics has always been about understanding your business and using that knowledge to make decisions. The problem comes when those decisions have to actually get carried out. All too often, good ideas come out of analytics, but fizzle into nothing when data actually needs to be put to work.

Reporting alone is necessary, but not sufficient. It doesn't actually drive the actions that move the needle. Modern companies can no longer just make data-driven decisions. They need to act on those decisions with data and do so automatically. This is operational analytics.

An example of Operational Analytics

Let's take an example of what it might look like in practice.

Imagine you work at a software company with a freemium model. Users can sign up for free and use the product up to a certain limit, at which point they then have to pay. You might use analytics in the form of a BI dashboard to track the number of signups, the percentage of users who convert to a paid account, and the effectiveness of sales reps in converting those customers to paid. You find that the sales reps who spend time personalizing the outreach to free users with information about their specific use case tend to over-perform.

Currently, these sales reps need to track down information across systems - Slack, Salesforce, and others - in order to get the full scoop before sending out a personalized, relevant email from Hubspot. This is where operational analytics can help.

With operational analytics, the same data that's feeding your BI dashboard can be automatically synced into Hubspot. For instance, Hubspot contact and account records can be enriched with with information like: whether or not the user has fully onboarded, the last login date, and the integrations that the user has set up. Now, sales reps don't need to track down information and spend time manually writing personalized emails. This data can be used to automate that outreach, leaving reps more time to help customers.

This isn't a hypothetical example. It's exactly how Retool used Hightouch for operational analytics. Once Retool began using analytics not just for reporting, but for action, they saw some pretty staggering results, including a 32% increase in reply rate on emails, as well as 500% increase in click rate and increased feature adoption.

Want to get started?

If you're ready to get started with operational analytics, we're happy to help. Feel free to create a shared Slack channel with us here at this link: https://api.hightouch.io/api/misc/shared_slack or schedule a call with us here: https://calendly.com/mwhittle5/meeting. There's a lot to consider with operational analytics, and some teams might not be ready for Hightouch just yet. That's okay; we aren't pushy and will do our best to help.