About SiteGPT

Problem Statement

Businesses generally struggle to provide instant, accurate, and personalised responses to customer inquiries 24/7. This challenge can lead to missed opportunities and diminished customer satisfaction.

​

Solution

SiteGPT is designed to enhance the customer service interactions. It is a platform that allows businesses to create personalised AI chatbots. These chatbots are uniquely tailored to the brand’s unique identity and are trained extensively on a company’s specific knowledge base. Whether it's websites, links, documents, or FAQs, SiteGPT’s AI chatbots can consume all provided information to deliver instantaneous and accurate answers to customer questions.

​

Value Proposition

SiteGPT directly addresses the pressing need for businesses to offer real-time, personalised customer support without the constraints of human resource availability. By leveraging AI chatbots that are specifically trained on your brand’s specific data, SiteGPT ensures that your customer support is not just responsive but also deeply informed and brand-aligned. This approach not only enhances customer experience but also positions your brand as responsive and attentive, leading to increased customer loyalty and satisfaction. With SiteGPT, businesses can efficiently scale their customer service operations, ensuring no customer query goes unanswered, anytime, anywhere.

​

​

Why SiteGPT?

I am the Co-Founder of SiteGPT and I usually deal with all the technological aspects of the product. I am picking up my product, SiteGPT for the Engagement and Retention project because I know everything about the product and I have access to all of its data like cancellations, retention cohorts, direct interactions with users etc. So, it make more sense for me to pick SiteGPT for my project.

​

Understanding SiteGPT

Core Value Prop

The main Core Value Prop of SiteGPT is the ability to “Answer the visitors queries instantly, 24/7 through AI powered chatbot”.

​

How do users experience the CVP

Users of SiteGPT experience this CVP every time their end users (website visitors) asks a question in the chatbot, and gets the answer from the chatbot instantly.

​

Natural frequency of the CVP

As per the CVP of SiteGPT, the natural frequency of CVP would be same as the frequency of the questions asked by our users’ website visitors.

Natural frequency of CVP = Multiple times in a day

​

Sub products of SiteGPT

There are multiple sub products of SiteGPT. They are mentioned below.

1. Chat History
2. Team Members
3. Q&A
4. Functions
5. Leads
6. Integrations

​

Natural frequency of sub products of SiteGPT

Best Engagement Framework for SiteGPT

Based on the above table, it is clearly evident that the best engagement framework for SiteGPT would be “Frequency” because no matter how many features the user uses, they experience the CVP more from more frequency only.

​

Core E&R Metrics

Actions that makes user an Active User

The following are the actions that the user has to perform for them to be called as an active user.

1. Create a chatbot
2. Embed the chatbot in their website
3. Real website visitor asks a question through the chatbot ⇒ At least 1 message in the chatbot

Natural Frequency of the product

The natural frequency of the product is “multiple times in a day”.

​

​

Customer Segmentation

Now, as we have defined who are the active users of our product, we now have to segment these active users so that we can run campaigns on some specific segments to make those users experience the CVP more.

The below are the ways in which we can segment all the active users.

1. Engagement Based - Casual / Core / Power

Our first customer segmentation would be to segment the users based on their engagement with the platform mainly into 3 buckets - Casual users, Core users and Power users.

Frequency is our selected engagement framework for SiteGPT. So, the engagement will be similar to the frequency of product usage.

Casual Users

Definition

• Casual users are those who have less number of messages from their visitors and use less number of features inside the product as they don't have much usage.

Features Valued

• These users are mainly concerned only with chatbot knowledge whether they have provided all the knowledge or not. The features that are valued by casual users are:
• • Adding/Managing Links
  • Adding/Managing Files
  • Adding Q&A

Natural Frequency

• Casual users will get < 10 messages per day

​

Core Users

Definition

• Core users are those who get decent of questions from their website visitors and they are concerned about chatbot knowledge and are looking to resolve any unresolved queries.

Features Valued

• These users are concerned with chatbot knowledge and also want to see the chatbot performance and resolve any unresolved queries by seeing the chat history. The features that are valued by casual users are:
• • Adding/Managing Links
  • Adding/Deleting Links
  • Chat History
  • Updating Q&A (based on user feedback and chatbot answers)

​

Natural Frequency

• Core users will get 10-25 messages per day.

​

Power Users

Definition

• Power users are those who get so many questions from their website visitors and they are concerned about the chatbot performance, try to get insights from the visitors questions and increase the resolution percentage.

Features Valued

• These users are concerned with the chatbot performance and mostly they have multiple members in their team looking to resolve any unresolved queries and also sometimes they might want to use some advanced features. The features that are valued by casual users are:
• • Chat History
  • Updating Q&A (based on user feedback and chatbot answers)
  • Members
  • Functions
  • Other Integrations
  • Leads

​

Natural Frequency

• Power users will get > 25 messages per day.

​

2. ICP/Persona Based

​Below are the 2 ICPs that we had prioritized during the Acquisition project.

​

Engagement Campaigns

Campaign 1: For Power Users

• Segmentation: Power users
• Goal:
• • Leverage integrations to maximize the frequency and quality of interactions between end users and the chatbot.
• Pitch/Content:
• • Expand Your Reach: Link SiteGPT with Slack to Maximize Engagement!.
• Offer:
• • Exclusive access to beta features for top integrators, plus special recognition in a community forum for innovative uses of integrations.
• Frequency and Timing:
• • Announce the campaign quarterly. Provide monthly check-ins, tips, and support to ensure users are effectively utilizing their integrations and to encourage them to try new integration features as they are released.
• Success Metrics:
• • Increase in the number of integrations implemented
  • Number of messages from these integrations
  • Participant feedback on ease of integration.
• Channel:
• • Targeted emails, and in-platform notifications.
• Explanation:
• • This campaign aims to deepen the use of SiteGPT among power users by harnessing their existing infrastructure, thereby increasing the natural interaction frequency. It also positions SiteGPT as a versatile tool that enhances their existing tech ecosystem, driving more frequent and meaningful engagements.

​

Campaign 2: For Core Users

• Segmentation:
• • Core Users
• Goal:
• • Increase the number of messages by leveraging specific chatbot features that increases visitor's engagement with the chatbot.
• Pitch/Content:
• • "Boost Your Chatbot's Impact: Discover and Master a New Feature Every Month!"
• Offer:
• • Receive bonuses or subscription discounts for actively using highlighted features each month.
• Frequency and Timing:
• • Each month highlights a different chatbot feature that increases the visitors engagement. Regular updates and prompts will be provided throughout the month to encourage adoption and mastery.
• Success Metrics:
• • Increase in message volume per featured function
  • User engagement with new features.
• Channel:
• • Targeted emails and direct messages through the admin interface.
• Explanation:
• • This campaign introduces core users to different chatbot features on a monthly basis, keeping the platform engaging and educational. By focusing on one feature at a time, users are more likely to integrate these features into their regular use, increasing their overall activity and improving their chatbot’s effectiveness. In this campaign, we specifically focus on features that improve visitors engagement thereby increasing the number of messages (CVP).

​

Campaign 3: For E-commerce Customers (ICP #2)

• Segmentation:
• • E-commerce Customers (ICP #2)
• Goal:
• • Enhance customer engagement and satisfaction by integrating real-time functionalities into their chatbots.
• Pitch/Content:
• • "Transform Your Online Store with Real-Time Chatbot Features! Enhance your customer service with immediate access to order statuses, loyalty points, and more directly through your chatbot."
• Offer:
• • Free setup consultation for the first month and a 10% discount on the premium plan for three months when they integrate real-time features.
• Frequency and Timing:
• • Launch the campaign at the start of a business quarter to capture strategic business planning cycles, with follow-up messages every two weeks to remind customers of the offer and provide success stories as examples.
• Success Metrics:
• • Number of e-commerce customers adopting the real-time features, increase in customer interactions via the chatbot, customer satisfaction ratings post-integration.
• Channel:
• • Email marketing, targeted ads on social media platforms frequented by e-commerce professionals, webinars focused on leveraging real-time data in customer interactions.
• Explanation:
• • This campaign targets e-commerce businesses looking to upgrade their customer interaction capabilities. By showcasing the direct benefits of real-time functionalities, such as improved customer service and enhanced user experience, the campaign aims to encourage adoption. The additional incentives like a setup consultation and discount on the premium plan further sweeten the deal, making it a compelling offer for e-commerce platforms ready to innovate.

​

Campaign 4: For Agencies

• Segmentation:
• • Agencies currently using SiteGPT
• Goal:
• • Encourage agencies to adopt the white-label offering to launch their own branded chatbot services.
• Pitch/Content:
• • "Elevate Your Agency: Offer Custom Branded Chatbots to Your Clients! Transform your service portfolio with our white-label chatbot solution—fully customizable to carry your branding and meet specific client needs."
• Offer:
• • Special introductory pricing with a 20% discount on the white-label package for the first six months, plus a free onboarding workshop.
• Frequency and Timing:
• • Announce the campaign at the beginning of a quarter to coincide with strategic planning periods, with follow-up communications bi-weekly to keep the offer top-of-mind and highlight success stories.
• Success Metrics:
• • Number of agencies signing up for the white-label offering, feedback on the onboarding process, and early adoption case studies.
• Channel:
• • Direct emails, dedicated webinars explaining the benefits and setup process, in-app notifications for current users, and personalized outreach via account managers.
• Explanation:
• • This campaign targets agencies that are already familiar with and see the value in SiteGPT. By offering them a white-label solution, we enable these agencies to leverage their existing knowledge and client relationships to provide additional value through customized chatbot services. The introductory discount and onboarding support are designed to lower barriers to entry, making the decision to expand their offerings with a white-label solution more attractive.

​

Campaign 5: For all Users

• Segmentation:
• • All users of SiteGPT
• Goal:
• • Drive user engagement and retention by providing actionable insights derived from their chatbot interactions.
• Pitch/Content:
• • "Unlock the Full Potential of Your Chatbot! Sign up now for our 'Monthly Performance Digest' and receive in-depth insights and trends directly from your chatbot's interactions."
• Offer:
• • Exclusive access to tips and best practices for chatbot optimization, plus a free custom report consultation for the first month after signing up.
• Frequency and Timing:
• • Users can sign up anytime, with the digest sent out during the first week of each month. Reminder emails or in-app notifications can be sent mid-month to encourage sign-ups.
• Success Metrics:
• • Number of sign-ups for the monthly digest, user satisfaction rates with the insights provided, and subsequent optimizations made to chatbots based on the insights.
• Channel:
• • Email campaigns, in-app notifications, webinars that showcase how to interpret and utilize the data effectively, and social media announcements.
• Explanation:
• • This campaign leverages the chat history export feature of SiteGPT to help users understand their chatbot's performance better. By providing a monthly digest of insights, users can see trends, pinpoint areas for improvement, and celebrate successes. The additional offer of a free consultation for custom report tailoring upon first signing up adds value, encouraging more users to engage with this service.

​

​

Retention Design

I am attaching the retention curve for SiteGPT, both Customer retention and Net MRR retention. Since SiteGPT is only 12 month old, I am plotting the average retention curves based on the monthly cohorts.

​

​

Based on this retention curve, it is very clear that SiteGPT has not achieved PMF. Both the retention curves are not flattening. In 1 year, almost all of the users are churning away. So, we are either not focusing on the right ICP or our positioning is too generic.

Based on the user calls, we understand that retention for customer support use case is far more than lead generation use case. So, the retention would look far better if we just focus on customer support.

And our main acquisition channels are word of mouth and SEO. Retention from people who come to us from WOM is a lot better than people who come from SEO. So, focusing on referrals might be a good way to increase retention.

​

​

Reasons of Churn

Users of SiteGPT churn due to various reasons. Some of the top reasons are:

• I am an agency, waiting for client to pay
• Not needed now
• Chatbot responses are not good
• Not the right product
• Less usage
• Not working properly
• Found better alternative tool
• Still exploring
• Want to pay via invoice
• Want features which are available only in Enterprise
• Too expensive (or) no money right now
• Building it in-house
• Training took too long
• Want whitelabel
• Interested in free tier

These are all the reasons we have gathered through various user calls and from in-app feedback in cancellation flows.

​

​

Let's divide all the top churn reasons into Voluntary and Involuntary reasons so that we can design resurrection campaigns accordingly.

Voluntary Churn

Voluntary churn occurs when customers consciously decide to stop using your service due to dissatisfaction or finding a better alternative.

1. Chatbot responses are not good: Indicates dissatisfaction with the AI’s performance or relevance of responses.
2. Not the right product: The product may not meet the customer's expectations or specific needs.
3. Less usage: Customers do not engage frequently with the product, indicating a lack of perceived value.
4. Found better alternative tool: Competitors may offer features or pricing that are more attractive.
5. Too expensive or no money right now: Price sensitivity or changes in financial circumstances.
6. Building it in-house: Decision to develop a similar tool internally.
7. Training took too long: Onboarding process may be perceived as too complex or time-consuming.
8. Want whitelabel: Desire for branding options not available in the standard offerings.
9. Interested in free tier: Only interested in free services, indicating a low commitment to paid features.

Involuntary Churn

Involuntary churn happens due to reasons outside the customer's immediate control, often related to payment issues or external circumstances.

1. I am an agency, waiting for client to pay: Cash flow issues caused by third-party payment delays.
2. Not needed now: Temporary or seasonal lack of need for the product.
3. Not working properly: Technical issues or bugs affecting user experience.
4. Want to pay via invoice: Payment process not aligning with customer preferences or business practices.
5. Want features which are available only in Enterprise: Accessibility of certain features limited by the product tier.

Negative Actions to Monitor

The below are some of the negative actions that we have to monitor so that we can prevent churn before it happens :

• Low Net Promoter Score (NPS): Reflects poor customer satisfaction and likelihood of not recommending your service.
• Low Customer Satisfaction Scores (CSAT): Direct measure of customer satisfaction with your service.
• Decreased frequency of logins or usage: Suggests declining engagement and potential disinterest.
• Negative feedback on specific features: Provides direct insights into what might be causing dissatisfaction.
• Removing team members: Removing team members from the chatbot might indicate the commitment from the user to decrease and could potentially lead to churn.
• Downgrade to lower plans: This could mean that the usage of our user is decreasing and they may be at risk for churn.

By classifying churn reasons and monitoring these key negative actions, you can tailor your engagement strategies to address the specific needs and concerns of different user segments, thus reducing churn and enhancing customer satisfaction. For instance, improving AI response quality could reduce churn for users dissatisfied with chatbot performance, while offering flexible payment options might help retain agencies waiting on client payments.

​

​

Resurrection Campaigns

Campaign 1: Improve Chatbot Responses

• Segmentation:
• • Users who churned citing "Chatbot responses are not good"
• Pitch/Content:
• • "Experience the New and Improved SiteGPT: Smarter Than Ever!"
• Offer:
• • Free access to the enhanced AI model for one month.
• Frequency and Timing:
• • Campaign launch as soon as improvements are deployed, with follow-up emails every two weeks during the first month.
• Success Metrics:
• • Re-engagement rate,
  • feedback on AI improvements,
  • subscription renewal rate post-trial.
• Explanation:
• • This campaign targets users who left due to poor AI performance. By showcasing significant improvements in AI capabilities and offering a free trial, users can directly experience the enhancements, potentially reversing their initial dissatisfaction.

Campaign 2: Right Product Fit

• Segmentation:
• • Users who churned because "Not the right product"
• Pitch/Content:
• • "Discover How SiteGPT Can Be Tailored to Your Needs!"
• Offer:
• • Personalized demo and consultation to showcase tailored use cases and features.
• Frequency and Timing:
• • Initial contact with an invitation for a demo, followed by periodic updates on new features every quarter.
• Success Metrics:
• • Number of demos completed,
  • rate of re-subscription post-demo,
  • user satisfaction scores
• Explanation:
• • This campaign aims to re-engage users by demonstrating the adaptability of SiteGPT to their specific needs, potentially correcting any misperceptions about the product’s capabilities.

Campaign 3: Competitive Edge

• Segmentation:
• • Users who churned because they "Found better alternative tool"
• Pitch/Content:
• • "See What You're Missing with SiteGPT - More Than Just a Chatbot!"
• Offer:
• • Comparison report between SiteGPT and competitor tools, showcasing unique features plus a 15% discount on their next subscription.
• Frequency and Timing:
• • Launch with the unveiling of new features, with follow-up reminders every month.
• Success Metrics:
• • Number of users returning,
  • engagement levels with the comparison report,
  • conversion rate from the offer.
• Explanation:
• • By providing a detailed comparison, this campaign highlights SiteGPT’s unique features that competitors lack, aiming to sway users back by emphasizing exclusive benefits.

Campaign 4: Economic Incentives

• Segmentation:
• • Users who churned due to "Too expensive or no money right now"
• Pitch/Content:
• • "Get Back to Growing Your Business with Special Pricing on SiteGPT!"
• Offer:
• • Reduced pricing or a flexible payment plan for a limited time.
• Frequency and Timing:
• • Reach out immediately after product updates or pricing changes; follow up quarterly.
• Success Metrics:
• • Re-engagement rate,
  • user feedback on pricing satisfaction,
  • retention rate after offer period
• Explanation:
• • This campaign addresses price sensitivity by offering special pricing or flexible payment terms, making it easier for users to justify the investment based on their current financial situation.

Campaign 5: Customization and Branding

• Segmentation:
• • Users who churned desiring "Want whitelabel"
• Pitch/Content:
• • "Brand It Your Way: Custom Chatbots with SiteGPT’s White Label Solutions!"
• Offer:
• • Introduction to the white-label options with a 20% discount for the first three months.
• Frequency and Timing:
• • Timed with feature rollouts relevant to customization, with periodic check-ins every two months.
• Success Metrics:
• • Number of sign-ups for the white-label option,
  • feedback on customization options,
  • retention rates post-discount period.
• Explanation:
• • Targeting users who require branding, this campaign promotes SiteGPT’s white-label capabilities, offering customization that aligns with their brand identity, potentially realigning the product with their expectations and needs.

​

Each campaign is designed based on user feedback and churn reasons, aiming to address the specific concerns and preferences highlighted by former users, thereby increasing the likelihood of re-engagement and sustained satisfaction.

​