Leveraging generalist support in the age of AI
To future-proof your organization, you need a generalist more than ever
2023 marks an era where AI offers more opportunities than ever for both employers and employees, but its true potential lies in its power to complement human capabilities rather than replace them. What are the capabilities of AI, and why is a generalist so important when integrating an AI strategy into your business?
With the arrival of generative AI as a powerful tool for white-collar jobs – software development, design, procurement, and medicine, to name a few – leaders are wondering how to best utilize it alongside their workforce without losing crucial insights provided by humans needed to drive their product forward. There is a balance between using AI efficiently and relying so heavily on AI that the product is an impersonal version of its original intention.
In early-stage startups, there is more uncertainty than process. While developing operational blueprints as you grow and scale, incorporating AI can feel like it will only compound that uncertainty. Unlike mature companies which may have a more straightforward use for AI and the data to back up their decisions, startups operate by constantly fine-tuning products and services to carve their path in the market. The challenge for startups is to embrace AI as an innovative tool for expanding capacity or streamlining systems, without becoming overly reliant.
The pair of a generalist and AI makes a perfect combination. Combining generalist and AI capabilities leads to stronger outcomes for ambiguous environments – where AI is capable of analyzing data, finding patterns, suggesting ideas, and managing schedules, a generalist has the creativity, insight, strategy, and adaptability that can make a product succeed when data alone will not be able to predict marketability and internal culture.
How do you combine a generalist and AI for the most impactful outcome?
Combine the power of data analytics with the nuance of cross-functional insight
Why This?
Using both data-driven insights and human-driven insights can launch a product more powerfully than relying on one perspective alone. Data-driven insights can provide a rough idea of what the current market trends are. But startups become successful by being disruptive – by seeing something beyond what is listed in market trends. An experienced generalist can work with AI to understand what the data is saying while having the creativity and insight to see what the numbers aren’t showing.
Here are ways in which both AI and a generalist can help with early-stage startups:
Data Analytics
AI can help analyze large datasets on market trends, consumer behaviors, and feedback sentiment.
A generalist can use this data to draw conclusions and make informed, complex decisions. Data analysis requires humans to interpret it and draw insights from the patterns.
Forecasting
AI can help predict trends and consumer behaviors based on patterns in previous data
A generalist will use forecasting and adaptability to ensure that unforeseen changes in the landscape are taken into account in a product's development.
Project Management
AI can streamline scheduling and suggest product timelines
A generalist can work cross-functionally, build relationships, and adjust timelines in an agile, dynamic setting
User Research
AI can identify patterns in user feedback and AB tests, and track behaviors during tests
A generalist has the empathy and emotional intelligence to understand the meaning behind user behaviors and identify pain points in a product
Customer Experience
AI can offer personalized product recommendations and handle customer interactions.
A generalist can build on these personalized templates and take into consideration the underlying ethics of personalization. Additionally, a generalist may see personalization features that do not yet exist and will likely not be recommended by AI.
Why Now?
The early stages of a company are the best time to establish a vision and having a well-informed vision makes a product unique and marketable. Having data-backed insights, and a smooth process to work cross-functionally will ensure a product’s success and the success of the company as a whole. Having the right generalists on the team (whether that is you as the founder or someone you've hired) who can be the cross-functional glue and consider the tools and processes where AI can be inserted into ops is a crucial part of building a future-proofed startup that's flexible enough to navigate these ever-changing times.
how We Do: Use Data to see the Bigger Picture
Use AI for a well-informed strategy
Tools 🛠️
AI can be a useful tool for data analytics, depending on what you need to achieve.
Some tools that could be used are ChatGPT LLM, ChatGPT’s code interpreter plugin, Tableau, and Google Cloud Auto ML.
ChatGPT LLM and the code interpreter plugin can help with generating code to analyze data or create visualizations, understanding which models to use, and documenting analyses.
Google Cloud’s AutoML builds and deploys custom ML models, depending on the needs of the user. It can help with custom model training, image classification, object detection, NLP, and model evaluation.
Tableau has generative AI capabilities and can provide automated analytics described in plain language, anticipate user questions, and suggest questions.
Rules (Process) 📝
It is important to have quality data and accurate analyses to develop market-driven decisions and develop a unique strategy to set your product apart.
Organizing and cleaning data takes a substantial amount of work but will affect the outcomes of your analysis, and poor data will lead to poor outcomes.
Using automated tools to organize and clean data to save time.
The type of analysis performed is just as important as having good data, especially if you are using ML and need to tune and evaluate a model to get accurate results.
People 🫶
The right person will be able to use these tools to better capture outcomes that align with the company’s needs, goals, culture, and competition.
These tools will only be as useful as the questions and the data they receive, and the output can only be useful if the user can find insights specific to the circumstance.
By leveraging AI and specific data insights, your team will be able to develop a more comprehensive, data-driven strategy.
Create an efficient, communicative culture
Tools 🛠️
To make communication easier and lower frustration, tools such as ChatGPT, Oracle, and PMOtto could be used. These tools are useful for automating project reports and addressing scheduling or progress issues across the organization.
Rules (Process) 📝
Virtual assistants such as PMOtto or Oracle’s assistant can provide status updates, update times, and track progress thereby finding efficiencies while reducing redundant reporting.
The virtual assistant can offer suggestions such as offering different time frames or extending projects based on changes in time allocation and project scope.
People 🫶
While AI can help with tracking tasks and making suggestions so the user does not have oversight, it cannot replace cross-functional communication, relationship building, adjustments due to unforeseen changes in the marketplace or staff, and the overall design of the original project’s implementation.
By implementing a culture that supports the adoption of AI while integrating it into your people strategy, you can create tighter, more streamlined communication loops.
It is important to carefully assess how, where, and when AI strategy intersects with people strategy by taking into account the task/team priorities (e.g. machine learning, analytics, UX research), as partnering with specialists will have better results. Ultimately, the main consideration is who should own the AI facilitation, depending on the task/team at hand.
Free up time for creative customer problem-solving
Tools 🛠️
AI tools can help with both market research and customer research.
Tools useful for market/customer research include anything that can help with sentiment analysis, survey analytics, video splicing, transcription, A/B testing, and behavior tracking.
One tool specific to user research is Userdoc, which generates user stories, user journeys, and personas.
Looppanel is another tool that automates aspects of the research process, such as interview transcriptions, notes, video clips, and data organization.
Rules (Process) 📝
There is a lot of human behavior that can be easily quantifiable and tracked through AI.
For example, to compare two interfaces, it is possible to track and analyze eye movement, mouse behavior, and time spent completing a task.
AI can create user personalizations, automate experimental procedures, and analyze feedback.
AI is also able to automatically generate KPIs, predict product success through historical data and market trends, and provide customer support.
Eliminating the need for the manual implementation of metrics, policies, and procedures can free up time for your team members to work more creatively as opposed to tactically.
People 🫶
Much of human behavior is not easily quantifiable, and quite nuanced and AI doesn't yet have us fully figured out.
Examining human behavior alone does not tell us the underlying intent of someone’s action, or how learnable a product is. Understanding the “why” of behaviors requires a person to look into why a behavior occurred.
Having empathy and insight is crucial to creating a product that users will feel engaged with, otherwise it will feel robotic.
Actually Actionable
The best way to create an excellent AI toolkit for a generalist depends on the company’s needs, but each team member’s role should be considered. It is also important to not rely too heavily on AI, or overload staff with too many tools. Tools that are easy to access and integrate into the user’s daily life will help them achieve what they are aiming for - to redirect their focus on larger problems and creative problem-solving. A few key questions worthy of consideration as you think through actionable ways to implement AI into your company-wide processes:
Do specific members of your team spend a lot of time working in data analytics?
Hold a meeting to audit and identify their biggest time-sinkers (data collection, data cleaning, data visualization) and invest in tools to address this specific area. Most people will find easily automatable areas of their work to be the biggest time-sinkers (1 Hour).
Review where they would like to invest more of their time instead and ensure they have the resources to do so (1 Hour).
Do they have to work across many diverse teams, such as hardware, software, marketing, UX, and data science?
Make it easier for them to keep track of the logistics, by integrating project management tools. There will be several other factors that they can maintain outside of scheduling (2 Hours).
Are jobs heavily reliant on building a cohesive culture across teams?
There are tools for many unique roles – selecting only the right ones, and no more will help them immensely.
The key to having the right tools is to identify the needs of each team annually. As a company grows and products and priorities change, pain points and focus areas will also change. Whereas a small company of ten could likely get by just fine without automated scheduling, a company of 500 is going to get bogged down with trying to keep everyone on the same page.
An annual tool audit with all key internal stakeholders will go a long way to ensure you avoid tool fatigue and onboard/offboard your tools effectively (1 Hour)
Before you go
AI is becoming a game changer in how people approach their careers. While it shows a lot of promise to automate many aspects of our roles, it still requires someone to build relationships, architect plans, and find creative solutions to the problems presented in data. With the right tools, a generalist will be able to free up time to focus on how to drive the company forward and how to make a truly unique and marketable product.
The oAT team will be taking a break from how We Do next week to celebrate the holidays, but will be back in your inbox on January 11th. Wishing you a wonderful holiday and a happy new year!
Writer: Cassandra
Interested in working with Cassandra through of All Trades to transform your product operations? Email founder@weofalltrades.com for more on how to bring her in as an embedded operator in your startup.