Marketing "Middleware" Put Simply

"Middleware," the term that everyone is talking about and no one understands...

My company, Improvado, raised our $8M Super Seed round this year— it was a big milestone for the company, as well as a another gratifying validation that we are solving some important problems.

We are in the marketing data "middleware" business. And over 65 mar-tech founders and investors agree that this is a space that is growing in size and importance.

What is middleware?

Middleware is software that connects two or more otherwise separated applications. Middleware software is a large and high growth segment with lots of companies, and hundreds of use cases.

Despite the market's size, everyone in it -- from marketers to investors -- could benefit from a clearer framework for how to think about middleware solutions.

The number of applications dedicated to marketing has exploded, up 20X over the last 10 years to over 7,000 platforms. The average marketer uses at least 12 different platforms (this includes platforms used to run ads, email automation, web analytics, social, search, SEO, CRM, CDPs and more) and this is likely an understatement.

‍

The growth of these disparate platforms has created a fragmented consumer experience. Today's consumer sees an average of 4,000 advertisements each day across an ever-exploding number of screens, devices, and formats.

The traditional marketing funnel is a thing of the past. The new customer journey is a patchwork of “experiences” that span advertising, marketing, word-of-mouth, and customer service.

With this backdrop, Marketing Middleware is the glue that brings this fragmentation together — whether to help make better data-driven decisions, or to unify the user experience across all of these touch points. In either case, Middleware is the solution to these problems.

While there are different types of Middleware that serve different verticals and team functions (for example, engineers utilize iPaaS), this article will focus on Middleware solutions for marketers and the differences between the various options.

The Two Types of Middleware

In general, you can split marketing middleware into two big groups - middleware for analytics and middleware for app synchronization, reflecting the reporting/optimization and user journey themes described.

1- Middleware for Analytics

Middleware for marketing analytics allows you to extract data and perform an analysis of the data. It reads data from many Apps.

An example of a tool that is middleware for analytics: Improvado.

2- Middleware for Data Synchronization

Middleware for data synchronization syncs data between two separate applications. It writes data into many apps for app synchronization or actions.

An example of a tool that is middleware for data synchronization: Zapier

Middleware for Analytics

See it in action

Let's start with a hypothetical example of how middleware can be used within marketing data analytics.

Mary works in the marketing department of a major e-commerce platform. Her team uses anywhere from 8 to 20 marketing tools and services within a given day. Here is a list of the marketing platforms where they are spending time and money every day:

Everyday, Mary analyzes marketing channel performance with two goals: optimizing budget, and reporting their work across departments and functions.

It's a complicated business process involving huge amounts of data that will be the basis for making critical and time-sensitive decisions.

This e-commerce business lives or dies by its marketing, so the stakes are high and mistakes have big implications.

That's where the need for a middleware solutions for analytics appears.

Mary needs to aggregate all this data from multiple SaaS platforms and join them together in order to get a holistic picture of every dollar that has been spent and how it impacts ROI and actual sales.

How it works

When it comes to middleware for marketing analytics, there are three different activities and functions. For short, we refer to them as ETV (Extract, Transform, Visualize). Together, this is the workflow of extracting and preparing data from SaaS applications for analysis. For each of these three steps, there is a software layer - meaning there are companies whose sole focus is to help marketers during each of these steps.

Extract - Data extraction layer
Transform - Data preparation layer
Visualize/Analyze - Visualization and analytics layer

In the image below, we've segmented software solutions for each stage of the middleware analytics process and also by solutions that require engineers or IT team members for implementation. Here at Improvado, we believe it's important to provide marketers the freedom to work with tools that allow them to integrate analyze their data without involving engineers.

The key players

When it comes to Middleware for Analytics, these software tools are doing the heavy lifting...

The best software for extracting your data (The “E” out of ETV”)

The extraction function is where data is pulled from multiple marketing applications for analysis.

The software that Extracts and Loads data connects via APIs to hundreds of SaaS applications and collects or pushes it into one or more databases.
‍

Companies that do this well: Improvado, Segment, Datorama, Singular

The best software for transforming your data

Those who have had the arduous task of integrating data from different platforms are all too familiar with the differences in naming conventions, definitions, and challenges applying equations across the aggregated data sets.

“Transformation” software normalizes data sets so that they can be analyzed together uniformly.

More advanced Transformation platforms are injecting machine learning to add a built-in layer of intelligence benefiting the analysis.
‍

Companies doing this well: Improvado, Trifacta, Paxata

The best software for visualizing your data

Dashboards, aka the Business Intelligence (“BI”) layer: after you prepare data you need to organize and visualize it in different ways.

Software companies solving this visualization problem provide interfaces with interactive charts to make analysis accessible and simpler for business users.

Companies doing this well: Tableau, Google Data Studio,

‍

Middleware for Data Synchronization

How it works

Now imagine you want to send data directly from one app to another. For example with a data synchronization tool, you would send a new lead from Facebook to Salesforce or from Salesforce to Jira or from Jira to Intercom. In reality, these are the most popular example use cases:

Identity resolution. CDP, DMP - Connecting a user across different platforms and devices, typically for cross-device frequency management or messaging
Data on-boarding - On-boarding User IDs such as email addresses into a SaaS app for targeting
SaaS app synchronization - Event-based synchronization like IFTTT (“if this then that”), to trigger an action based on a previous action in a separate platform
Tag management - Sync events from your website or mobile app with other SaaS apps, based on pixels.

In the image below, we've segmented software solutions for each use case of data synchronization.

The key players

The best software for identity resolution

These platforms collect user-level data across multiple applications and devices; they'll often serve as a central hub for user-level data, and allow the development of different user profiles and segments that can inform action within other platforms.

These platforms are often charged with reconciling cookies, mobile IDs, or other user IDs.

‍
Companies doing this well: LiveRamp, Mparticle, Segment

The best software for data on-boarding

Data on-boarding is a way to append additional data to your user profile database before loading it into marketing or media buying platforms.

For example, if you want to connect email addresses with offline data from publicly available databases such as home ownership, or credit card purchase data from Mastercard, an on-boarder would help you do this.

Companies doing this well: LiveRamp, Mparticle

The best software for app synchronization

Synchronization is a way to send information from one platform to another, to trigger a certain action.

For example, if a person completes a poll in Survey Monkey, a marketer may want to send her an email from MailChimp.

Typically you use these platforms to synchronize applications or for cross-channel script.

‍
Companies doing this well: Zapier, IFTTT, Workato

The best software for tag management

Tag Management software is used for managing different pixels intended to track, or trigger activity data to be pushed from your mobile app or website to other applications.

There can be some complex logic involved in how these pixels speak to other applications, and tag managers help marketers managed the instructions without the use of software code — i.e. tag managers will launch javascript on websites, and automatically send events to SaaS applications based on rules established in the interface.

Companies doing this well: Google Tag Manager, Tilium, Segment

Evolving Your Mastery of Middleware

Before you navigate or take action in the middleware space, it's important to understand it — I hope this article has been helpful for framing the different use cases, categories and sub-categories of functionality, and identifying some of the key players across the spectrum.

Keep an eye out for future writings where I'll attempt to give people a framework for thinking through what's important should they find themselves contemplating investing, partnering, or looking for middleware solutions for their companies.