Thalox Customer Engagement Score: Your Key to Smarter Segmentation

Ditch outdated lead scoring methods and focus on predictive engagement. thalox’s dynamic customer engagement score is the future of HubSpot Marketing.


thalox in 3 steps

thalox was built for HubSpot. To perform the native integration with HubSpot, follow the simple HubSpot OAuth process: log in, choose your workspace, and connect the app. It’s really that easy! Upon connecting thalox to HubSpot, it will quickly extract and analyse your contact data as well as their behaviour based on all digital historical activities using the latest AI capabilities.

Predictions power thalox. Running a prediction is simple and quick. First, name your prediction, so you can keep track of it alongside the campaigns you’re using in HubSpot. Next, our machine learning algorithms will be trained on your data and analyse your entire database, assigning each contact a customer engagement score between 0,00 and 1,00.  The higher this value is, the more likely a contact will engage with your next email marketing campaign. We call this our thalox engagement score (TES)

After a complete database analysis, you’ll be delivered 3 segmented contact lists based on their thalox engagement score.
  • Dormant Potential Segment
  • High Engagement Segment
  • Inactive or Indecisive Segment

In just 1 click, you can push these segments into HubSpot! They will automatically appear among your HubSpot lists, ready to be used in Marketing campaigns. We provide recommendations of best practices for each segment. Clean out inactive contacts (or re-engage them), tailor messages to each segment, and watch your emails convert better!

What are the 3 thalox segments?


High Engagement Segment

Contacts that are very engaged with your brand and will likely engage with your e-mail marketing initiatives. Continue with your current strategy and test new messaging.

Dormant Potential Segment

Contacts with the most potential to exploit and move into the high-engagement segment. Concentrate your marketing efforts here.

Inactive or Indecisive Segment

Contacts that are inactive or indecisive about if they want to interact with your brand. Re-engage or clean these contacts.

General FAQ’s

  • Where is my data stored?

    All your data is stored capsuled from other customer data on a AWS server in Frankfurt.


  • Is thalox for marketers connecting to other Marketing Automation Systems like HubSpot?

    It is on our roadmap. Please let us know which system you like to see: Contact us!

  • What is the difference of HubSpot Score and thalox Score?

    HubSpot Score is based on the input a marketers is putting into as a parameter. Thalox Score is the result of a machine learning algorithm with no human interaction. Our system is learning by itself. For more information please read the following article.

  • How many contacts do I need to get good results?

    We recommend > 500 contacts which at least 10 activities per contact

  • Where can I find help?

    There are a bunch of articles in our knowledge base. We recommend to starting with the “Getting Started” section.

  • Does thalox provide professional services?

    thalox is a software company. But in some cases we help customers leveraging thalox for marketers and training them how to use thalox for marketers the best way. Our agency partners are here to help – if you have any specific questions please fill out this form and contact us.

  • Can I get marketing support from thalox?

    Our founders and most of our employees have a strong marketing background but we are no agency. What we do is interpreting your data when we jump on a call with you. At some Enterprise customers we also do some marketing support services but normally we recommend an agency from our network. Please reach out to us to see how we can help you.


  • What type of model is created to generate the scores?
  • Which activities are consumed in the ML model?

    Currently we are taking the Email activity from your HubSpot system.

  • How many fields are included in the prediction model?

    There is no individual answer to that since our model is learning over time. That means the model could start with > 20 fields and over time the model is taking > 50 fields when it detects those are contributing to the model.

  • Can I choose the properties I want to score?

    No – our model automatically takes the fields that show a high correlation with the customer engagement. This usually also changes from prediction to prediction, as our algorithm is able to learn.

  • Are my custom fields from HubSpot reflected in the algorithm?

    Absolutely yes.

  • Why do you show a density chart and a pie chart?

    If you love data the density chart gives you much more insights into your data. You will hopefully see over time that the peaks are going more into the direction of a wave.

  • What is the difference between HubSpot unengaged contacts and the “Inactive” segment from thalox?

    The definition from HubSpot is: A contact will be categorized as unengaged if they meet one of the following criteria: Never engaged with a marketing email from you and hasn’t engaged with the last 11 emails you’ve sent them. Previously engaged with one of your marketing emails but haven’t engaged with the last 16 emails you’ve sent them.

    Thalox “Inactive or Indecisive Segments” predicts the likelihood someone is engaging with your next email.

  • What is happening with my data when I do not want to prolong with the service?

    No – only of you want to prolong with our service

  • Is there a limit on the number of predictions per month?

    Currently we encourage you to do a weekly prediction to see your individual results, so that you can learn with your data. Currently there is no limitation in the Business and Enterprise package.

  • Do you have a sandbox to test first after going live?

    Currently there is no Sandbox environment since the systems needs real data. But you can cancel your existing environment (Trial) at any time.

  • Should I delete my contacts with no activities?

    There is no general recommendation for this question, since there could be many reasons for that. But we recommend having a closer look into those contacts.

  • How often do you distribute new features?

    Currently we are running 3 weeks sprints cycles which not always have new features included. Normally there are new small features deployed every quarter.

  • Does bad data quality influence the TES?

    Data quality has a significant influence on AI results. AI algorithms rely on large volumes of high-quality data to make accurate predictions and decisions. Poor data quality, such as missing values, inconsistencies, inaccuracies, or biases, can lead to flawed AI outcomes.

Elevate Customer Engagement Scoring with thalox