Sage runs marketing automation using a central model whereby all Eloqua campaign work is completed by the Global Marketing Automation Team. A daily challenge we faced in the Marketing Automation team was the volume of questions received from the stakeholders (Marketers) within the business. Typical questions that are received range from best practice marketing automation, Eloqua functionality, our processes, SLAs, Templates, Lead Scoring, etc. What we noticed was that many of the questions were very similar and we were spending a lot of time answering the same questions over and over; time which could have been better spent automating and executing campaigns.


The goal of this project was to decrease the time spent answering questions and increase time spent executing campaigns. We also wanted to make the process easy for our stakeholders to find the correct answers to their questions very quickly so they could focus on the bigger tasks with higher priority at Sage;better campaigns and increased ROI.


We wanted to do this with existing resources that were already available to us; the answer was Eloqua.


What did we build? A chatbot which contains all the answers asked on a regular basis and more.



Meet the ‘Demand Marketing Chatbot’.



To put it simply the chatbot was created using the following:


  • An Eloqua Landing Page
  • A Form
  • Google Sheets
  • TableTop (An open-source JavaScript library)


The Google sheet which we called the ‘Chatbot Brain’ contains all of the questions and answers which the chatbot expects to receive.



The functionality

When a user asks a question to the chatbot, the data from the ‘Chatbot Brain’ is ‘crawled’ by the Tabletop.js script. A bit of custom JavaScript then matches the question to an answer using a ‘fuzzy match’. The relevant answer is then displayed on screen to the user as a response. If a suitable answer can’t be found in the ‘Chatbot Brain’, the question that was asked gets recorded (to another Google Sheet). This document can be  monitored and those questions and answers can then be added to the ‘Chatbot Brain’ in case they’re asked again in the future.


As it’s built on an Eloqua landing page the tool is hugely scalable by sharing the URL and has been rolled out on a Global basis for Marketing teams within Sage. A link to the Chatbot has been added to all the various internal tools and resources that the stakeholders use so they always have easy access to it.


This is the first chatbot type tool that can be used within the Marketing team at Sage, giving quick and important answers to its users.


The assets which are needed to build it are so simple but effective which means they can be used again for any type of chatbot, including for external marketing activity. It has been suggested that a similar tool could be used externally for customers to get instant answers about our products.


The Results

In its first three months, the stats include:


  • 400+ landing page views
  • 90+ visitors
  • 4.5 average views per visitor
  • 340 questions asked
  • 111 new answers added to the chatbot brain
  • An 80% reduction in questions directly to the team

This last statistic is hugely important to us as the team have spent less time answering questions from stakeholders and stakeholders now have a place to go to get fast answers. This means that Marketers at Sage can focus on creating and implementing revenue generating, cost-effective and innovative marketing campaigns.


As a Global Marketing Automation team, building campaigns for Marketers across the business, our customers are Marketers at Sage. B2B: Engage and Enrich helped us to decide how we can improve the experience they have with working with us as a team. We wanted to increase engagement by creating a tool which was user-friendly and had a look and feel that they are familiar with. The functionality is easy for them and encourages them to use the tool more frequently and the content is tailored to their needs.


During the implementation and roll-out of the Chatbot we continuously checked on the landing page insights to ensure that the tool was being used. It has helpful to see that we had returning visitors from the landing page analysis overview which meant we could see that the tool was helpful. Without these insights, it would have been difficult to see if the tool was a success, however, having over 90 unique visitors in 3 months proved that it is working very well. The B2B: Insight for Reporters course showed us how we could use reports to make decisions about campaigns and the next steps to increase effectiveness.


Huge thanks to kevincassidy313 for his support throughout this project!