Strategy Series: New Technology Adoption [AI]

Written: August 16, 2024

Always strategically thinking

The great thing about working in the consulting space is that you often solve problems in very different spaces. For better or worse you hear different situations and play out how you would solve those situations.

We're talking Digital Business Transformation. If you want more resources on it I highly recommend Nigel Vaz's book.

I was lucky enough to have had leadership support to attend Transformation Leadership Accelerator.

Disclaimer: none of this is rocket science, there's no special sauce here, no industry secrets. It's the application of business concepts that bleed over into the engineering space.

We should [AI]

I was talking with some colleagues about work and it had come up that the business they work for should do [AI], which can be a very complicated thing to answer.

In the realm of [AI], different businesses are at different points in their adoption and we're all waiting to see how the inflection point takes off as businesses report on value in quarterly earnings. So far, that value has not been achieved and it's leading to many tech companies needing to cut workforce until that investment turns the corner.

While commercial, government, and academic spaces are different, there's a lot of strategic overlap. Much of the adoption of [AI] is similar to adoption of other technologies over time, so let's take a look at that.

What is the landscape? Understand the space your business is in

This is a competitive analysis to understand what kind of company you are, who is your competition, what are your competition doing to differentiate themselves in the space?

It's understanding what you as a business are also doing in the space. Maybe y'all haven't started, or some groups have.

Are you on the bleeding edge pivoting weekly or daily to stay ahead of the competition? Are you just beginning and behind the competition? How much does your org see this specific tech transforming it? This needs to be included in your vision and how hard your plan needs to achieve.

In business, increasing revenue means bringing a new offering to market, selling a current offering to a larger portion of the market, or operating more efficiently. When performing the competitive analysis these opportunities need to be considered. If the company is investing in [AI] just because without a vision, then likely it will become a wasted investment. This is happening in much of the space; businesses are investing without measurement as part of the plan to understand if they should continue or not.

Where do you want to go? Set the strategic vision

Armed with the competitive analysis, the understanding of where the business is, and the direction we want to go; a strategic vision and plan on how to get there can be put in place. A "we also do what our competition does/me too" plan could have a different timeline than one that's meant to meet and go beyond the competition.

Communicate out what that plan is, the objectives, the milestones. Help people understand that it's not all known and that some things will be learned on the go, perhaps there will be mistakes, pivoting will be needed, and we need their help to be successful.

OKRs and KPIs need to be communicated and agreed upon to understand if the investment is having the impact it needs to. It's accountability, but less about data to fire people for underperforming and more about knowing if the org needs to help pivot or that they need to better support the groups.

Create that vision, it needs to come from high up within the org to have the commitment and investment needed if you want it to succeed.

Make sure to answer:

• What kind of actions do you want to take to differentiate the business?
• What kind of cost is associated with that? How do you budget for that?
• How fast do you lean into that vision?
• What's the business model to make money based on these new capabilities?
• How do you plan on communicating your new capabilities to the market?

Who leads the charge? Groups have differentiated context

• In commercial space it could mean software engineering, data analysis, visual experience design, logistics, operations, production, staffing; the lists are vast. The kinds of groups will vary across verticals of business, but may overlap. Fast food and big box stores both have logistics teams. Services industries and product businesses both have customer support groups.
• In government groups could mean different departments, which can wildly be different across areas such as social support and revenue/taxation to politics and military applications.
• In academics it could mean the different colleges within a university: fine arts, mathematics, computer science, business, heath sciences, physics, architecture, etc as well as the operational core groups supporting the organization.

What's critical here is that one can't stand up a centralized [AI] group and expect that group to solve all of the business groups' issues. Only each of those groups know the problems and the opportunities to potentially apply the tech to.

An organization needs to stand up and fund a center of excellence that can dive into that space and help with the application of that technology to each of the business groups. They need to enable champions to understand the space and bring that understanding to the other business units. This is critical especially for the technology spaces because many of the business groups will not include large sets of technical software engineers that can explore this themselves. Each group will bring their talent and thought leadership within the org to say, we can transform our org with this tech by applying it in this way, but I need help in doing it.

In this case the champions of the new tech (in this case [AI]) need to have:

• The pulse of the space and competition
• Understand ethical implications to promote responsible implementations that promote equality
• Sustainability understanding (like the enormous amount of electricity needed by [AI] tools) to leave the world for our children
• The time/budget to experiment
• Relationships with leaders in other business groups to understand the use cases
• The accountability and capability to continually share across the org

How to get there, identify the milestones

1. Crawl

   • Setup the initial champion group
   • Ensure initial champions have access to tools
   • Communicate the plan, secure the budget
   • Create the short term roadmap, pencil in longer term aspects
   • Implement OKRs and KPI to understand how well the goals are being achieved
   • Identify leaders in each of the business groups to support the roll out
   • Get pilot programs off the ground, celebrate their successes
   • Identify necessary initial talent and acquire talent

2. Walk

   • Create and rollout training plans, plan for burst needs in talent
   • Get more business groups involved
   • Raise awareness
   • Solidify the longer term roadmap
   • Review KPIs, ensure meeting business objectives for value
   • Increase investment to support additional value generation
   • Update the competitive analysis, pivot roadmap and plans as needed
   • Ensure larger talent pools have access to tools

3. Run

   • Have collaborative retrospectives, share the output, help teams pivot through the changes to stay on track.
   • Update the competitive analysis, pivot roadmap and plans as needed
   • Ensure all teams have access to tools
   • Continual training of existing talent

Technology is easy, people are the hardest part... Change management

1. People fear the unknown, train them, give them access, set their expectations

   • Training is necessary, make sure the vision and rollout plan has these aspects incorporated.
   • Access to technology is huge, once people work with a technology they understand the potential and how it can help achieve great new things.
   • Talk to people about how this technology isn't meant to replace them but to assist them and allow them to do more.

2. Expect resistance

   • Understand the influencers with in the org (both for positive adoption and negative resistance).
   • Support the adopter champions, but also listen to those who resist, winning them over is key to turning the ship without an anchor holding the org back.

3. Build trust

   • Commit to keeping people in the know, and following through on it.
     • Have newsletters, lunch and learns, retrospectives, and quarterly reports on how that vision is being achieved.
   • Without trust an org may eventually integrate a transformative technology, but it will be adopted much slower, creating lost opportunity cost from the time not leveraged.

4. The right support

   • Junior talent is great. They tend to have:
     • A more open mind on new technology
     • Are ripe for continued learning
     • They don't have organizational bias and are less jaded
     • They typically are less expensive which is great for investment opportunities like [AI]
   • Throwing a bunch of junior talent at a technology without the right guidance is a recipe for swirl, leading to lost investment across all contexts.
     • Make sure they've got the right support/leadership which has bandwidth to support the new investment.

Engineering: making the technology happen

Startup companies are disruptive because they are able to pivot quickly, something that larger/long established businesses have trouble doing. Whatever the technology is that helps businesses operate in the digital space, it needs to be adopted in a manner that can be nimble, that can pivot quickly or the competition will beat you with feature speed to market every time.

Here's a few engineering aspects to consider:

• Consider build vs buy. With the rapidly changing space of [AI], it may be better to buy (SaaS) for the moment until things settle down. Also it can let the org understand if this is the right investment. Add build considerations into your digital roadmap.
• Avoid mainframes that have significant costs and timelines to update.
• Architect in manners that can be adjusted quickly.
• Consider domain layers to insulate third party providers.
• Leverage modern software practices and make sure automation is leveraged when possible, especially for quality.
• Consider scalability, resilience, security, and other considerations of the [well architected framework](https://aws.amazon.com/architecture/well-architected).
• You may not have in-house talent that knows the ins and outs, consider contracting/partners that can help set you up in a manner to own the longer term solution.
• Data. There's so much of [AI] (machine learning, generative [AI] or otherwise) that depends on solid data. If your data lakes/warehouses/pipelines aren't solid all the [AI] tech in the world can only do so much.

There's also product aspects that go beyond engineering to consider:

• Look at agile instead of waterfall implementations for speed to course correct.
• Have open feedback cycles on the tech; if it doesn't work, adoption speed will decrease. Consider a product manager to help drive the work towards value.
• Adopt a product mindset, that it's not just a project and done (Capex), but a living entity that will need continual support (Opex) after roll out.

Application and research

There's two main spaces which can apply across the three contexts above:

1. Application of the current [AI] concepts to groups

   Assuming the business isn't one of the main [AI] as a service providers (Google, Meta, OpenAI, Microsoft, AWS, etc), lots will be invested in applying what's currently known in the space to business problems. This could be big data analysis or generative [AI] applications (imaging, text, RAG, Agents, etc) to solve business problems in ways competitors haven't conceived of. Blue Ocean opportunities may apply here, but the [AI] space is booming with startups carving out new areas of the market or outpacing others in those markets.

2. Research into new [AI] concepts

   The main [AI] services providers are throwing millions/billions to fund research, it will be hard to outpace their advancement. By understanding what aspects of research are progressing, it may be possible to focus on other areas that are not receiving as much funding to make advancements in adjacent spaces. Researching adjacent areas also allows Blue Ocean opportunities. This would likely require inputs from areas including but not limited to software engineering, computer science, data, mathematics, and business. Many innovations come when cross-disciplinary groups combine to create new ideas/applications of concepts.

Only for [AI]?

All of this applies across technology whether we are talking [AI], [crypto], [blockchain], [social media], [printing press]. The same solutioning will apply to the next technology as well. It's why all the references to [AI] have been in brackets, so when the next transformative technology is developed this similar set of concerns can be leveraged.

Summary

Implementing a new technology to help transform your business requires more than an agreement that "we should [AI]".

It requires strategic vision, support from high leadership, planning, and continual support of your teams to make sure it provides value to the business and your customers.

Perhaps you or your org hasn't started this journey, perhaps you are in it and it's not going well.

Whatever the case, if you or your org is looking to implement a digital transformative aspect, reach out! Perhaps we can collaborate to create a more concrete plan or pivot an existing one for your business.