Launching and Scaling Your AI Business - 7 Tips To Scale

AI, AI businesses and anything related to 4IR is attracting a lot of support, focus and funding in the current markets.? If you offer AI to a corporate, the response is an unequivocal “We want one!”.? There is strong political and budgetary support.?

We are in the early stages of this market and there is no doubt that AI will become ubiquitous and be an integral part of product-led technology businesses going forward.? How AI is deployed and leveraged will provide a significant advantage to businesses that get it right.? The timing to launch and scale product-led AI businesses has never been better, but AI businesses come with their own set of challenges.? In this article, we look at ways to make the process of creating and launching your AI business a bit easier.

Three Steps To AI Engagement

To unpack the market and processes, we simplify the AI engagement into three steps

• On-ramps? - Gather, structure, clean and prepare the data
• Highway - Analyse the data to accelerate insights
• Off-ramps? - Turn the insights into action

Self-driving cars are a great example, the on-ramps are where the data is created.? The sensors, instruments and GPS gather the data, location, speed, proximity of other vehicles, speed limits, direction, possible threats and maps this against where the vehicle needs to end up, the destination. The AI algorithm is the highway, this analyses the data to provide insights to decide what to do to get the car to its destination without breaking the rules and avoiding accidents.? The off-ramps are where an action is created. The action determines the success of the process.? Turn the wheel, indicate, brake, accelerate etc.? Success is getting the car to its destination by the best route, on time and accident free.?

Closed AI Systems

Closed AI systems are statistics “If I do x y times the output will probably be z”

The self-driving car processes happen in near real-time.? This is possible because it’s a relatively closed system.? The sensor and input data is presented to the algorithm ready for consumption and the data is processed and acted on immediately because the vehicle controls are under the direct control of the ‘robot’.? There are a finite number of inputs and variables, roads are relatively structured environments.? Self-driving cars are a growing and highly competitive sector.

Open AI Systems

Open AI systems are calculus “How many times do I have to do x for the result to be z”

The complexity multiplies in a business environment.? Poor, unstructured, distributed, inaccessible, incomplete and incompatible data sets combine with AI algorithms that are often new and have not been validated in this specific environment.? These datasets are then applied in business environments where we rely on people to achieve outcomes we are often not sure are optimal.? The destination is usually to be better without understanding exactly what this means.? The complexity explodes as we extend solutions across the Economic Value Chain, where the opportunities to optimise are relatively unlimited.? For a market sector that is driven by ‘science’ and data a large amount of subjectivity creeps in around the predicted business outcomes.? Evidence based solutions get faster business support.? In more open B2B AI systems the on-ramps, the highway and the off-ramps each can provide significant challenges.

On-ramps & off-ramps Drive Scale

The on-ramps (how you ingest the data) and your off-ramps (how the data is deployed for the customer to act on), frame how well you can serve your market and scale your business.? Many AI businesses start off by building the highway, the tools that accelerate insights and then look at the on and off-ramps.? Building a solution where the off-ramps require organisational behaviour change or where the on-ramps require months of data preparation may offer high value but often do not make the customers ‘to do’ list.?

Build Effort Required for Scale

On-ramps - 40%

Highway - 20%

Off-ramps - 40%

Scalable AI businesses are more about how to leverage AI than they are about building AI, scale is about the on and off-ramps you serve.

Here are Seven Tips to look out for

#1 Solve in Sequence

“If you don’t know where you are going you may end up someplace else” Yogi Berra

Focus on the off-ramps first.? The off-ramps are where the value is created inside the customer’s business.? The off-ramps will inform both the data blueprint and the product architecture. Your offering moves from hard-to-sell to easy-to-buy based on your off-ramps.

#2 Seek Accessible Datasets

“No matter how easy you make it for the customer, it’s not easy enough”

Customers will often lead you to datasets that they understand, these are often not easily accessible.? In post-covid markets, customers have low appetite for large complex projects that consume a lot of internal resources. The less you ask of the customer, the higher up the ‘to do’ list you will be. Datasets outside the firewall are often easier to access.? To shorten the Time To Value for the customer, to make your solution ‘easy to buy’, focus on more easily accessible datasets to start.

#3 Seek To Solve Existing Pain

“How you sell is more important than what you sell”

Accessing a new budget in post covid markets is challenging.? To shorten sales cycles, seek to identify and solve Customer Problem Statements where the customer has already identified a need.? If the customer has to seek a new budget or define a new business process to leverage the insights, then you are adding yet another variable to your project where success is far from guaranteed.? Start off by making sure your initial insights are built to solve existing Customer Problem Statements.

#4 Do Enough

“Confused customers don’t buy”

Many new AI businesses work hard to deliver highly extensible, low latency solutions.? Providing real time analytics is only relevant if the insights are acted on immediately, not the next day, week or month.? Low latency solutions are expensive to build, deploy and maintain.? The optimal latency is based on the timeframes the customer is able to act in.?

Data scientists and developers get excited about what is possible when data is combined with contemporary technology and add more and more features and possible outcomes.? If you solve a problem the customer values they will engage, more features or lower latency is not necessarily more value. Increased complexity increases confusion, not value.? Provide enough insights to create action that drives value.? Any time invested in making the processes simple and easy for the customer is rarely wasted.

#5 Build What You Can Sell, Don’t Try To Sell What You Can Build

?“If you build it, they won’t come”

AI businesses usually start off with a “this is what we do” sales pitch as they have spent a huge amount of time and energy building an amazing highway. Data scientists tend to think ‘it’s just data, it’s all the same to me’.? The product becomes sales-ready when the sales process shifts to “this is exactly how it will work in your environment, these are the benefits you will derive and this is how we integrate with the way you work”.? Selling what you have built leads to “Who is my customer” and “We can serve many sectors” scale challenges.? The less work the customer has to do to receive value, the more sellable your product is.? Does the customer think your on-ramps, highway, and off-ramps are ‘built for me’?? Don’t try to sell what you can build, build what you can sell.

#6 Create High Customer Empathy

“If you think you are in a niche, go lower”

B2B AI solutions seek to improve business processes.? It is important to align the benefits you offer directly with the customer's needs.? The currencies a customer values are framed by the sector they trade in, the customers they serve and the maturity, life stage and size and goals of their business.? Focusing on a niche eliminates competition, makes distribution easier, increases customer value, fixes the off-ramps, helps optimise the highway, reduces the on-ramps and enables you to build functioning sales funnels.? If you think you are in a niche, go lower.

#7 Action Creates Value

“We’re on the road to nowhere, come on inside” Talking Heads

Value is only created by action.? The data and the solution has no value if it is not acted on, irrespective of the time and effort spent on building and deploying it.? Most AI solutions can provide more data and insights than most corporates are able to act on in many months.? More insights do not equate to more value.? It is the job of the AI business to seek and deploy actionable insights, insights that fit into the customers existing business processes and offer quick value.? The value of your business is based on the value the customer derives, not the insights delivered.? As the domain expert, the problem is yours.? If the customer is unable to act on the insights, offer a solution they can act on.? The ‘last mile’ is usually the hardest, turning insights into action, but this is where the value is derived.

Summary

High growth happens when you are able to shift customer value from calculus to statistics.

Focus on a niche to serve existing Customer Problem Statements to deliver a solution using currencies the customer values.? Your business becomes scale-ready once you have solved for the on and off-ramps and offer a short Time to Value.?

Shifting your B2B software business from an Open AI system to a Closed AI system moves your product from ‘hard-to-sell’ to ‘easy-to-buy’.?

Sw7 is Africa's First Virtual Accelerator Platform For B2B SaaS Product-Led Technology Business Funders and Founders.? If you are going to market or looking to get funded and scale, you can get your free scale-readiness roadmap at www.Sw7.co with additional awesome support for #AWStartups with #AWSActivate.