Author: Mallory Lee

MadKudu was founded in 2015 by Sam Levan, Francis Brero, and Paul Cothenet with one important mission in mind: empowering modern marketers by making them more intelligent. What they have built gives marketing operations superpowers. Over time, their platform has come together with a variety of value-add features to increase intelligence for the entire sales and marketing process. (think signals) 

I’m not JUST talking about the kind of intelligence that tells you where John Smith works – I’m talking about enough statistical intelligence and signals to create a world where sales never questions marketing again. 

Their reach extends all the way from using a Clearbit integration to enrich your lead records, to providing you with ideal product + website conversion paths, to amazing lead scoring, and plug-and-play sales playbooks. Their platform really shines for PLG use cases with the ability to fully ingest product usage events and surface engagement signals from places like Github. 

My Experience: 

I used MadKudu while at Nylas, and specifically we used Segment to pipe in our product events. MadKudu digested the data, compared to our customer base, and found the best possible scoring algorithm for us to use in GTM. Along with a direct feed from the marketing automation platform, they can support the need to understand how marketing engagement data should be mixed in with product usage, firmographic information, and more – resulting in a data-backed MQL definition. 

Their team (and the tool) will point you to the “happy path” of ideal events to take place that result in the highest likelihood of converting to a customer. Just staring at a feed of product usage events isn’t all that helpful, so using a tool such as MadKudu to bring that usage journey to life is important. 

The caveat here is that your company has to have enough volume of data to make it predictive. That’s not always easy in the early days. Nylas had 1,000+ customers to base our scores on, which was very valuable. When we launched a new model, it took some time to build the relevant data set.

From my perspective as a customer, the initial work of running the machine learning to arrive at scoring models was mostly completed during our proof of concept – which was a big value add. We can’t say if that will be the MadKudu process forever, because it may scale better as part of the official implementation after the sale. But it has always been clear that the team has some serious analytical firepower on staff. The caliber of their services team was important to me because there are some complex ways to consider making changes to models; it’s not something I would feel comfortable doing alone currently. Reviews from other users share the same sentiment.

In recent months, and for much of 2024, MadKudu has added myriad integrations to bring additional signals into their data set. The most exciting ones that stood out were Navattic, RB2B, Clay, and LinkedIn. Signals from any of their integrated data providers can be factored into a prospects’ engagement score. In addition to using a qualification threshold, users can also create a more rules-based approach to qualifying if a certain signal is captured. (such as a former champion getting a new job) Another point of input from reviewers is that the Salesforce integration is currently stronger than the Hubspot integration for MadKudu; FYI.

Room for Improvement:

The team at MadKudu has the incredible challenge of making some serious analytics accessible and simple enough for users to make changes on their own. There is still progress to be made in this area, because the process of drilling into the simulated outcomes of various scoring model changes is not something the average user can easily do in a self-serve way.

Something buyers should consider when looking to adopt tools that incorporate signals: What’s the data licensing and retention policy? For example, the Clearbit data offered by MadKudu is invaluable but cannot be exported and saved in Salesforce. It’s only shown on the iframe and it’s up to you to build some automations in MadKudu to surface that as a high score and then go flag it or manually change the Salesforce record. 

What's Coming Next?

After a recent conversation with Chief Product Officer Francis Brero, it’s clear that MadKudu has firmly entered the fray of “signal selling platforms”, but with a unique twist. Francis emphasized their intention to act as a sort of “CDP for Sales” – able to digest all signals and make sense of them – regardless of which system of action a company uses. (think: Salesforce, Gong, Outreach, Salesloft, etc.) At the same time, they have deep experience in using machine learning to put together a true MQL definition that can incorporate these signals and spur the right sales actions. 

Francis also mentioned that MadKudu plans to add AI to the platform in a way that allows users to better navigate the complex tasks of modeling, without learning a complex UI. Imagine chatting to the system and asking it to evaluate the impact of changing your scoring model. For current MadKudu users, this will go a long way to improve the ease of use. 

When you boil it all down, I still think MadKudu is a platform for marketers, because the best marketers focus on enabling sales execution. 

*Note that future looking statements and roadmap previews are never a promise of feature work or delivery. You should only make decisions about buying technology based on what is available to you right now. 

Disclosures: 

• Mallory was a MadKudu customer in 2023
• Mallory has appeared on the PLG Confessions podcast
• There has been no compensation associated with this independent review, but we hope it's helpful to MadKudu and their buyers