SmartSend and AI recommended content
Improved the effectiveness of internal and external content sharing among customers by an average of 70% when compared to traditional methods.
Project overview
SmartSend delivers real-time AI-powered content recommendations to your sales reps in their flow of work, tailored to your buyers' interests and interactions.
SmartSend empowers enablement and marketing teams with real-time content insights to enhance buyer engagement, resulting in accelerated sales cycles and increased revenue.
My contributions
This strategic initiative unified teams across R&D, mobilizing resources from every department. Despite our compact two-person design team, we orchestrated the complete product lifecycle—from initial opportunity analysis and user research through iterative design, validation testing, and a successful beta rollout.
Glossary
FE: Front-end
BE: Back-end
FS: Full-stack
QA: Quality assurance
AI: Artificial intelligence
CMS: Content management system
Launch date: February 6, 2024
x1 Senior product designer (that’s me 👋🏻)
x1 Product design manager
x3 Engineering managers
x6 FE engineers
x3 FS engineers
x2 BE engineer
x5 QA engineers
Scope planning
Our R&D team collaborated to define the initiative's scope, establishing clear objectives and developing user stories that aligned with our strategic goals. We utilized the role-goal-benefit framework to help the team write effective user stories and define feature requirements.
- Role: The type of user who will interact with the feature
- Goal: The action or functionality the user wants to accomplish with the software
- Benefit: The reason the user wants to accomplish the goal
Design ideation
We strategically allocated work across our squads based on our agreed-upon scope and detailed user stories. Our early start on design exploration allowed us to rapidly prototype and validate solutions, delivering a working end-to-end version that enabled development to begin on core functionality.
Uploading files
A key phase of SmartSend was expanding Spekit’s CMS capabilities. This meant allowing for users to upload their own files into the product. We started by supporting PDF and PowerPoint (PPTX) files, marking the first phase of our broader vision to support diverse content types and integrations within Spekit.
File management
To help users instantly distinguish between different content types in Spekit, we crafted distinct visual identifiers through unique color schemes and icons. The scalable architecture of the card design allowed us to accommodate seamless integration of new content types as the platform evolved.
Sharing files internally and externally
SmartSend also introduced secure content sharing through generated links, enabling sales teams to distribute materials both inside and outside the organization. This streamlined sharing capability helped sales representatives deliver content to customers and prospects while providing valuable engagement analytics.
Previewing an external file
We also gave users the option to preview the interface their recipients will experience before sharing content. The simplified, polished interface ensured frictionless content consumption for external viewers.
Tracking engagement
SmartSend's share link tracking revolutionized how users monitored content engagement. Real-time notifications in the Chrome extension instantly alert you when recipients view your shared content. Whether through notification alerts or the analytics dashboard, you can access detailed insights about how your shared content is performing.
Component restructure
We rebuilt core design system components to enhance visual consistency and streamline functionality across the platform. Comprehensive developer documentation accelerated implementation time and reduced technical bottlenecks.
Read my Medium article to see how I did it →
Prototyping the end-to-end experience
During development, I created an end-to-end prototype that visualized the complete user journey, ensuring alignment across cross-functional teams and providing developers with clear specifications for intended behaviors and interactions. This interactive demonstration proved instrumental in building stakeholder understanding and maintaining consistent product vision throughout implementation.
External beta program
I led the research strategy for our beta program, designing a comprehensive feedback framework to capture both qualitative and quantitative insights throughout the product lifecycle. The program successfully engaged 12 active customers from an initial pool of 37 invitees, who participated in a structured two-month evaluation process including kick-off sessions, real-time testing with asynchronous feedback, and bi-weekly review meetings. This systematic approach informed our post-launch enhancement roadmap with actionable user insights.
Pre-beta metrics
A new way of gathering feedback
I proactively established a centralized feedback system in Airtable during the beta program, streamlining our research operations by systematically cataloging user insights and interview data. This structured approach enhanced our data analysis capabilities and created a single source of truth for the team.
Common feedback themes
Ease of use
"Honestly, it's super easy to use and access."
"Very easy user interface."
"I like how easy it is to update information since our environment is constantly evolving."
Post-beta metrics
AI recommended content
While running the SmartSend beta program, our team developed AI-powered content recommendations for the Spekit Chrome extension, prioritizing Gmail and Salesforce integrations before expanding to general domains. By meticulously mapping the user flow and page-scraping algorithms, we gained crucial insights into how to surface the most relevant content within users' workflows. This visual approach to understanding our recommendation engine enabled more strategic decision-making and refined targeting.
.jpg)
Finding inspiration
I conducted a comprehensive competitive analysis of leading platforms—including Facebook, Pinterest, Instagram, and X—to identify proven recommendation patterns and user engagement models. These insights informed our design decisions and helped establish best practices for implementing AI-driven content suggestions that align with users' mental models.
Recommendations in the flow of work
We strategically integrated the recommendation feature into our extension's home page interface, optimizing existing space to enhance content discovery. By prioritizing recommendations in the visual hierarchy, we ensured users could immediately access contextually relevant information upon launch, maximizing efficiency while maintaining a clean, intuitive experience.
Feedback mechanisms
To continuously improve our recommendation algorithm, we implemented a multi-tiered feedback system that enabled users to refine the model through both general input and granular content-level responses. This strategic approach to user feedback collection enhanced our machine learning model while empowering users to actively shape their personalized content experience.
Launch day buzz 🚀
The launch of SmartSend represented a significant milestone for Spekit, mobilizing a comprehensive cross-functional initiative that spanned multiple departments. Our coordinated go-to-market strategy, anchored by a strategic social media campaign, generated immediate market traction and substantial user engagement, validating the high anticipation surrounding this feature release.
"Designing products around the people that will use it the most, embedding their feedback, and a commitment to making their lives easier... YES!"
"I was impressed with the work flow. I especially like your vision to make the process like Instagram in terms of being graphically pleasing, low friction to use and customized, relevant streams."
"All of the content at a sales reps fingers without having to search in multiple systems is a game changer. More time spent with the prospect than having to find content. Congrats Melanie and Spekit team on this awesome launch!! Let's go!"
