(Notes: All opinions are my own)

Introduction

The direction of travel of the global technology industry and its related analytics applications is going to be guided by several underlying trends in the next 5 years.

This brief highlights the ones most likely to affect industry dynamics and identifies the corresponding relevant market opportunity.

1) Auto-ML algorithms will be employed for an increasingly varied range of use cases

Definition

Auto-ML, or low code machine learning (ML), is the process of automating machine learning pipelines deployed towards a business application, such as demand-prediction, customer segmentation, and fraud detection.

The case for change

Current machine learning pipelines are multi-faceted and require intensive computing resources to conduct data-heavy activities, such as data labelling, cleaning, model training and hyperparameter tuning, and model deployment.

While data science human capital required to problem-solve and develop the applications that deliver the required uplift in business performance is critical, many steps of current ML processes are going to be subject to automation, thanks to mainly three converging factors:

a) Cross-industry implementation of academically tested algorithms which display a clear industry use-case fit and have been proved to deliver significant performance improvement (neural networks in computer vision, regression-based methods in sales prediction, clustering techniques in customer marketing)

b) Use-case repeatability which ensures a large segment of applications can leverage transfer learning from one implementation to the next, with marginal changes to the algorithms and solution infrastructure

c) Continuous focus on open-source software implementations, which displays an ever-growing collaborative push towards automating the most intensive and data-heavy ML pipeline components, leading to positive spill-over effects along the automation spectrum.

Risks and takeaways

In the most optimistic of scenarios, the auto-ML phenomenon will likely follow a Pareto Distribution, with 60–80% of common ML business use cases being subject to medium-to high degrees of pipeline automation.

Depending on industry and use case specificity, custom deployments will still be required, while less innovative firms will leverage Auto-ML to tackle more basic and prevalent use cases which will focus on quick-wins for their businesses.

2) The commoditization of cloud technologies will accelerate competitive dynamics in an expanding market

Definition

Cloud infrastructure has become essential to deploy world-class enterprise applications at scale and with greater cost efficiency. In this sense, cloud-resources have become the “technological commodity” and infrastructure upon which many applications rely to establish market presence and match modern performance expectations.

Current skewness of market share distribution will likely evolve in the medium-term due to two major effects described in the next paragraph.

The case for change

The market for cloud infrastructure displays the characteristics of an oligopolistic, “winner-take-most” industry, where the top 3 players serve the majority of demand.

The competitive development of the cloud industry will be interesting to watch and may evolve substantially depending on two important factors:

a) Levels of offer differentiation which each player will be able to secure by leveraging proprietary IP and strategic assets in key service areas (GCP in ML/AI via to their Tensorflow offering, Azure via their wide-ranging product and database suite, AWS via the sheer scale of their current ecosystem.)

Specialization around specific assets will gradually start tilting the competitive balance and grant expanding rewards in a constantly growing market, now that the full suite of basic services is converging and will no longer play a key differentiating role.

b) Ease of integration with renowned SaaS offerings, as the continuous push for application development via Agile and DevOps practices demands easier and faster integration requirements to enable developers more possibilities to build via a development approach further dominated by microservices.

For cloud players, staying on top of the latest frameworks and industry demands will be vital to continue to expand and solidify their customer base. This will likely be not a matter of choice, but a matter of speed, in which players able to integrate and implement faster will secure stronger returns.

Risks and takeaways

In this environment, business strategy will be strongly interlinked with cloud strategy and cloud-agnostic SaaS solutions and services will be of growing importance.

As the battle for cloud market share continues, service gaps will also represent key innovation areas around which other players will build fresh and novel value propositions.

3) The growing activity of Asia-Pacific private equity financing into the tech sector will have profound cultural repercussions

Definition

The future state of the analytics & technology landscape is being more and more driven by Asian influences, primarily China-led, as the nest of AI-based innovation has shifted from being located primarily in Silicon Valley [1].

The rise in Eastern intellectual capital development and innovation has being followed by private equity capital efforts, which are bound to resonate worldwide in the tech industry.

The case for change

China’s rise to AI power has been incredibly interesting to follow, and its influence across the analytics and tech world is bound for further accelerations in the years to come.

A major catalyst of such evolution is the rise of Chinese private equity financing [2][3], which is well supported by national interests determined to capture further technological advancements.

Capital flows by themselves, though, will not be sufficient to achieve further technological dominance, but will instead work in combination with the two following factors which will likely enable China to establish further AI and Tech industry dominance, thus exerting stronger cultural influence as a whole.

a) Human and intellectual capital development, spurred in part by a new generation of ex-Silicon Valley entrepreneurs which have launched massively successful tech companies, integrating elements of the US tech industry with Chinese-led application architecture.

b) Aggressive go-to-market strategies [4], thanks to massive capital investments, pervasive trade-relationships and deep pockets which will allow companies to secure stable market positioning and capture further growth.

Risks and takeaways

The analytics and technology landscape is a geopolitical matter rather than merely a scientific one. There is high uncertainty around what future market regulation and national policy are going to look like, but cultural influence exerted over the West by AI innovation coming from the East is likely to rise.

4) Digital markets’ regulation in the years to come will likely affect M&A activity in digital, with follow-on impacts on overall levels of innovation in a range of applied tech fields

Definition

Technology regulation will become more pervasive across industries. Future regulation is likely to impact a wide range of sectors, beyond legacy incumbents in telecommunications and Big Tech players. A strong regulatory push is going to impact technology providers across cloud, cybersecurity, and fintech industries.

Future levels of regulation will affect the degree of industry concentration and competitive dynamics amongst firms, and with those determine firm-level R&D strategies. Interesting new jobs and fields of research and academia are likely to start playing an important part in this evolving environment.

The case for change

New public attention towards regulation of digital markets, especially following a pivotal year in 2020 [5], will be likely followed by legislation which will impact future antitrust cases.

While regulation of the technology sector is unlikely to dramatically change industry concentration in the short term, some interesting impact should follow around the below areas of activity:

a) Greater Antitrust scrutiny on M&A activity in tech will impact R&D strategies at the firm level. When estimating likelihood of relative success in the analytics and software market and deciding whether to develop long-term innovation or target lucrative exits, higher regulatory risks are going to impact innovation investment decisions.

b) Increased regulation on data sharing and use will focus on identifying algorithmic bias and establishing transparent data ethics, with interesting new jobs and application potential. In highly regulated industries, machine learning applications will be scrutinized for human-bias and ethic design, and not merely optimized for business performance. Finding the trade-off between these two will likely require combined efforts on the part of several stakeholders, and its degree of success will also reflect the increased public awareness of such topics.

Risks and takeaways

Regulation around data and business practices in the technology industry will focus on areas that are deemed as business critical for the major players involved, often trying to influence those practices that are embedded in their business model and have proven to be key to establishing their competitive advantage.

Data-sharing regulation and open data standards will be critical in ensuring transparent conduct on the part of all stakeholders to make sure the public sector, end users and tech players can exploit the full potential of data.

At the same time, striking a balance between regulating harmful practices while maintaining sustained innovation levels will be key in allowing organic growth in the technology industry.

Summary

Hope you found this article useful. Feel free to share your opinion by commenting below or by getting in touch via my profile.

Thanks for reading!

Sources:

1. AI Superpowers: China, Silicon Valley, and the New World Order, Kai-Fu Lee, 2019

2. Number of private equity transactions in China from 2009 to 2019

3. Value of private equity deals worldwide in 2019, by target country

4. To Beat China on Tech, Biden Will Have to Learn from It

5. House Report Attacks Tech Giants as Monopolists That Stifle Competition

Access my free Data Science resource checklist here