How To Implement a Winning AI and Automation Strategy 2023
I’ve been hanging out in chatrooms, talking to some automation experts, completing AI certificates on #LinkedInlearning, and reading great content that I received about creating a winning AI and Automation strategy.
Your AI and Automation Strategy
Let’s talk about #Automation…
One of my favourite automation experts is Ed Axe. He runs an agency called Axe Automation.
This is what he recently wrote about startups and SMEs needing automation.
One of the most common questions he is constantly being asked:
“Ed, I know I need to start using automation, but I have no idea what tasks to start automating”
The hardest part about getting started is gaining clarity about what needs to be automated the most.
To help you with this…
In today’s edition, I’ll be explaining the simple process I use to help my clients identify the best possible automation opportunities in their business.
Let’s dive in:
Step #1: Time Audit
The first step is to completely audit your time and your team’s time for a week.
You need to get clarity around what tasks are taking up the most amount of your time each week.
There are 2 main ways to do this:
1. Manually track your time
You can use a simple worksheet or notebook to manually track your days for an entire week.
Or…
2. Use software to track your time
You can use an app like Toggl Track to measure your time for an entire week.
Both ways work.
Step #2: Identify Automation Opportunities
Once you’ve tracked your time for a week and identified what takes up the most time…
Now it’s time to identify automation opportunities.
From all the tasks that you do, you’ll need to see which you can automate.
Keep in mind:
The biggest opportunities are the tasks that take up the most amount of your time.
Automating these tasks will have the most impact on your operations and will bring you the highest ROI.
So how do you identify which ones can be automated by software tools?
By breaking your tasks down into 2 categories:
Category 1: Digitally triggered (using software)
Category 2: Human triggered (using creativity, for example)
The digitally triggered tasks are the ones you’ll be able to automate.
Step #3: Record the Tools you Use
Once you’ve identified the digitally triggered tasks…
Write down the tools used during these tasks.
This will give you clarity on every single software you’re using, paired with the task.
Step #4: Categorize your Tasks
From here, you’ll want to start categorizing each of these tasks.
Categorize your tasks into 4 different sections:
Section 1: Tasks consuming 10+ hours/week
Section 2: Tasks consuming 5+ hours/week
Section 3: Tasks consuming 1+ hour/week
Section 4: Other tasks
This way, you’ll know exactly what the best automation opportunities are for your business.
Start with the tasks that take up the most amount of time, then work down from there.
Final thoughts:
After doing this entire process…
You’ll know:
1. How you spend your time each week
2. What tasks consume the most of your time
3. Which of those tasks can be automated by software
4. The software tools you use to complete them
5. How much time each of them consumes
Now, all that’s left to do is:
1. Build a flowchart of the processes you need to automate
2. Build the automation for it
3. Test it and launch it!
But if you didn’t get clarity to start, you wouldn’t know what to automate in the first place.
Use this process to help identify exactly what you can automate.
Even executing one automation could save you 10-20 hours per week, depending on the task.
That’s how powerful it can be!”
How good is that as an automation playbook to get started?!
Your AI and Automation Strategy
Let’s talk about #AI…
I still sense resistance from most Startup Founders I engage with about the need to use AI tools to gain efficiency in their daily operations, as well as integrate AI into every product or solution they build and sell.
Artificial Intelligence (AI) is today’s most important enabling technology. Leading startups, scale-ups and enterprises are using AI today to reimagine consumer experiences and business processes – unlocking revenue growth and cost savings at the expense of their competitors. How can you take advantage?
Embracing AI can seem daunting. How do I develop an AI strategy? Can I afford AI? Do I need an in-house team? Which algorithms should I explore?
I’ve put together a framework that might help some founders take the first step towards AI.
MMC Ventures is a research-led venture capital firm that has backed over 60 early-stage, high-growth technology companies since 2000. MMC’s dedicated research team provides the Firm with a deep and differentiated understanding of emerging technologies and sector dynamics to identify attractive investment opportunities. In other words, when it comes to new tech like AI, they talk the talk and walk the walk.
Their AI Playbook is an accessible, step-by-step guide to taking advantage of AI. They explain, without jargon, best practices in six core competencies for AI: strategy; people; data; development; production; and regulation & ethics. In a hurry? Every chapter includes a one-page summary and checklist of actions.
Whether you: are a founder, executive, information worker or developer; work at a startup, scale-up or enterprise; and have a budget of £500 or £5m – this Playbook will catalyse your journey.
This is what you’ll find inside.
Contents
3 Introduction
4 Summary
12 Chapter 1: Strategy
How to:
• Identify and prioritise problems for AI to solve
• Evaluate AI deployment strategies – from third-party APIs to in-house teams
• Plan budgets and timescales for AI projects
• Build buy-in for AI and mitigate cultural concerns
22 Chapter 2: People
How to:
• Understand the different roles in an AI team
• Structure an AI team according to your objectives
• Source, evaluate, attract and retain AI talent
32 Chapter 3: Data
How to:
• Develop a data strategy for AI
• accelerate data acquisition
• Structure, secure and provide data
• Develop a high-quality data set
• Understand and minimise bias
44 Chapter 4: Development
How to:
• Understand the advantages and limitations of different development approaches
• Select a development strategy
• Choose hardware for AI development
• Address problem domains with suitable AI techniques
• Evaluate the strengths and limitations of popular deep-learning frameworks
66 Chapter 5: Production
How to:
• optimise your research and development activity
• Select a hosting environment
• Transition development systems to live use
• Measure and monitor system accuracy
• Develop a robust quality assurance process
• Implement an effective maintenance programme
76 Chapter 6: Regulation & Ethics
How to:
• comply with GDPR data handling requirements
• Verify that automated systems meet regulatory stipulations
• explore different approaches to ‘explainability’
• Apply a framework for ethical data use
I hope you’ll take the time to download the complete playbook here.
As always I welcome feedback and comments. Your experience might be different to mine and your feedback or comments will surely help another Founder gain clarity.
I do this out of gratitude for what investing in and working with Startups has given me over the last 30 years.
Please pay it forward…
About the author
James Spurway is an Angel Investor, Mentor, Advisor, Speaker, former Commercial Pilot, and Author who specialises in raising debt and equity capital.
He strives to model diversity, equity, and inclusion in the founders he agrees to invest and work with.
He has paused his angel investing activity to focus on raising his first US$ 50M venture capital fund, which will invest in startups that can accelerate the achievement of net zero emissions.
James spent the past 33 years living in Hong Kong, Vietnam, Germany, Switzerland, Monaco, the USA, Thailand, the Philippines, Singapore, and Australia, his country of birth.
In that time, he started 10 businesses, exited from seven, shut down two, and kept one. He has invested in a total of 50 startups since 1993. In the period from 2003 until 2020, he had two successful exits from Unicorn startups. His average ROI sits at around 5X to date.