Empowering Human-Led AI Conversations: Accelerating Product-Market Fit for a B2B Chat Pioneer

The Client: Revolutionizing the AI chat space

Revolutionizing Business Communication: Our client is on a mission to transform the way businesses interact with their customers. By harnessing the power of AI, they aim to enhance human connections rather than replace them. At this stage, our focus was on empowering small to medium-sized businesses with a more efficient communication process, while preserving their unique and personal touch.

‍

The Mission: Interim Product Manager

Our team comprised two engineers and one designer with a close partnership with the Chief Product Officer (CPO) and Chief Technology Officer (CTO). As the company was in its early startup phase, all team members were new to the organization. At this stage, there were no established processes, ceremonies, or standards in place, presenting a unique challenge for our team to navigate.

As the first Product Managers for the company, our mission was to establish essential product processes and frameworks, jump-start the engineering team, and ensure that delivery velocity and cadences met the rigorous standards required for an early-stage startup.

We were entrusted with laying the groundwork to achieve Product-Market Fit, a critical milestone for the company's growth and success.

‍

Our Main Quest: Achieving High-Velocity Delivery

Launch: Onboarded Very Fast

Upon joining, we encountered a lack of standardized Jira practices, development processes, and product frameworks. To address this, we initiated a kick-off workshop with the developers and designer during our first week. Our goal was to establish our ways of working, team commitments, and cadences, ensuring we built a self-led, accountable team driven by a missionary mindset rather than a mercenary approach.

First Deliverables and Results

1. Unified Vision: Conducted a Kick-Off Session with the developers and designer, leading to well-defined expectations and concrete steps for establishing cadences and delivery processes that support high-velocity delivery.
2. Streamlined Collaboration: Implemented Jira workflows and best practices in line with our team commitments, resulting in more efficient daily collaboration and delivery within the product and tech team.
3. Predictable and high-commitment delivery: When we joined there was no clear indication of what is possible to be delivered over a certain timeframe which created friction and stress among the cross-functional team members. We implemented ticket scoring and capacity budget planning which had the impact that developers were protected from unsustainable expectations and business stakeholders had comfort with predictable delivery cadences.

‍

Ready… Set… GO: Build on the foundation, lightning fast!

Our strategy required a delicate balance between delivering essential features that serve as the entry point for our users and introducing innovative, impactful features that differentiate us in the market. This approach was guided by the Kano model, ensuring that we not only meet but exceed customer expectations.

📚 Kano Model: The Kano Model categorizes product features into five types: Basic (must-have), Performance (satisfaction scales with performance), Excitement (unexpected delights), Indifferent (neutral impact), and Reverse (dissatisfiers). It helps prioritize features based on their impact on customer satisfaction. You can check our video about using the Kano Model. [LINK]

‍

Initiative 1 - Voice Innovation Powered by AI

Problem:

• Voice Messages had been established as a core use case for our client’s customers and their customers.
• It became important for our client’s customers to be able to make use of more innovative voice solutions that would allow them to be received as authentic and personable by the customers they are talking to.
• This is not always possible, as they might be in loud spaces or not have the opportunity to make voice messages of a high-quality.

What has been done:

• With integrations in ElevenLabs, we created the opportunity for our client’s customers to create deep fake versions of their voices that create voice notes from written messages.
• Our client’s customers were also able to opt for a preselected AI voice, which would create voice notes from written messages as well.
• All these voice capabilities could speak in over 29 languages.

‍

Initiative 2 - Streamlined content extraction powered by AI

Problem:

• Businesses would typically have resources that have all the typical information included that would answer most of the questions they get asked by their customers.
• However the entire resource or document would not be relevant to the specific question a customer could be asking, and it would take time for our client’s customer to extract the necessary information from resources and documentation to help their customers.

What has been done:

• Using the AI functionalities from third-party providers, we could suggest contextual responses to our client’s customers through our client’s app, with references to the appropriate pages and media in the appropriate documentation and resources.

Smaller, but important features were also developed and delivered during our time with this client:

• Full Front-End tracking set in place using Mixpanel: When we joined there was no front-end tracking in place, which made it difficult to make data-driven decisions. Within less than a week all front-end tracking for existing had been put in place, and front-end tracking was made part of our definition of done for all future features.
• Scheduling integration using Calendly and Google Calendar API: A typical job-to-be-done for a B2B customer in the chat space is setting up appointments, with a mix of Calendly and Google Calendar API integrations, we solved this job in less than a week.
• Two-way translations in both voice and text: All the actors in the client ecosystem could communicate with each other in their language of choice and the app translated those messages into the receiver’s language of choice.
• Jira backlog down to 40 tickets from 200+: As Jira best practices were not able to be put in place by the time we joined, it was difficult to manage priorities and delivery with the number of tickets in Jira, this was cleaned up and reprioritized to help the team work on the most impactful issues, and not simply deliver low impact features faster.
• Led and implemented the entire Data Privacy Hub and Data Processing Agreements: Being a B2B messaging app, privacy is of utmost importance. When we joined privacy by design was not included in the development processes. We implemented the MVP of the data privacy protection agreements and processes for our client.

‍

Laying the Groundwork for Product Market Fit: Our Side Quest

During the early stages of a startup, one of the most important sources of direction to ensure you reach product market fit is talking to as many of your target customers as possible to understand their needs and goals as well as possible.

• The conversations were happening, but they were not being documented, which had the impact that those insights were lost:
  • We created a sales and customer dashboard on Notion where all conversations should be stored and where data on industry type, company size, and met and unmet needs needed to be captured.
  • Understanding which customer segment had sufficiently met needs as compared to their unmet needs would ensure focus on which segments are ready to initiate marketing efforts.
  • Understanding which customer segments had insufficient met needs as compared to their unmet needs would ensure focus on which unmet needs are to be prioritized for delivery.

• Before having the front-end tracking in place, we were blind to features’ usage and whether they were solving any needs for client’s customers:
  • We implemented Mixpanel tracking and dashboards for all the existing features and ensured that front-end tracking was part of all features’ definitions done.
  • Being able to see which features are not being utilized by our customers helped to shed light on previous product decisions that might not be serving the business anymore, and which were candidates to be removed from code maintenance or improved.
  • Being able to see certain features being used more often than others also provided us with valuable insights into the real value our client’s customers are getting from certain jobs to help drive our direction further and get closer to creating a stronger unique selling proposition (USP).

‍

Mission Achievements: Delivered Outcomes

💡 Implemented 12 major features alongside maintenance and improvement issues over 16 weeks.

💡 Set up the team for sustainable and high-velocity delivery, implemented Jira best practices, and eliminated 160 redundant Jira tickets for more focused delivery.

💡 Implemented Product Discovery boards and templates linked to continuous delivery practices.