In 2024, chief technology officers (CTOs) at consumer companies grappled with the challenges—and embraced the opportunities—of the evolving technology landscape. Learning to use and extract value from AI, modernizing legacy tech systems, meeting ever higher consumer expectations, and complying with digital-trust legislation were all top of mind for these leaders.
In the year to come, these issues will become only more pressing. To stay ahead of the challenges, consumer CTOs—or the top tech leaders in retail and consumer-packaged-goods organizations, which could also include the chief information officer or chief information and digital officer—can prioritize eight tech-forward imperatives we’ve developed based on our research and analysis of hundreds of large-scale digital and AI transformations. These imperatives can not only spur innovation and accelerate growth but also boost efficiency and refocus operations to deliver customer-centric results in an increasingly competitive landscape.
1. Elevate tech teams from support staff to core function
Many consumer leaders continue to think of their tech departments as support functions that are responsible for executing choices made by other functions. But the most successful consumer companies make tech an integral part of their strategy. To maximize the value of tech investments, consumer organizations can involve tech leaders in setting the broader business agenda and hold them accountable for business outcomes.
To do this, consumer CTOs should embed tech capabilities (including software, AI, and enterprise data) into joint business-tech teams. Functionally, this means that tech teams take part in business strategy meetings and, ultimately, have a say in how tech investments are determined. CTOs can also update the organization’s incentive structure to merge tech and business outcomes and metrics. For example, a tech team that builds the company’s e-commerce site would share responsibility for sales volume, since a better checkout experience can lead to higher sales.
Rather than treating AI as solely the tech team’s responsibility, one large consumer staples company required its product innovation, marketing, and supply chain departments to collaborate with tech teams to experiment with and incorporate AI within their functions. Each tech investment the tech team helped facilitate was connected to a business outcome—for example: using a particular machine learning (ML) solution to plan promotions also led to increased sales. Further, the business invested in a proprietary ML platform, which now powers AI models used across 80 percent of its global business. One of the platform’s uses was to launch a tool that helps parents choose the right diaper size for their child. By giving the tech team a seat at the proverbial table—rather than relegating it as an order taker—the business can get more from its AI investments in the long run.
2. Empower cross-functional AI teams to be problem solvers, not just orchestrators
Leading consumer players have already established centers of excellence (COEs) that help make their businesses more agile. But a COE does not have to be limited to acting as a project manager that deals with challenges after they arise. Instead, a COE can influence overall AI and gen AI strategy, partner with business units or functions to prioritize domains and use cases, and develop risk mitigation plans, including guardrails for AI adoption.
For that to happen, CTOs can empower a COE to educate other functions about AI’s potential and proactively offer solutions rather than simply fulfill requests. A COE should also work with functional experts to identify the metrics to measure AI’s impact across the business.
On matters that affect more than one domain (for example, architecture choices), a COE can gather the appropriate experts across domains to make joint decisions (for example, cloud architects, platform engineers, and an API expert may all need to weigh in on cloud-native architecture). An AI COE will also have visibility into how a solution for one team—say, a chat interface—can be modified for another.
One big-box retailer was already aware of the power of COEs. The company had multiple AI COEs for various functions, which helped each function capture the benefits of adopting AI quickly while bypassing red tape. However, the approach wasn’t scalable. These disjointed efforts prompted organizational leaders to build a central AI COE. In addition to standardizing the retailer’s AI work across business units, the AI COE gained visibility to identify the shared challenges across groups adopting and scaling AI use cases. Despite these gains, the retailer still hadn’t unlocked the full potential of its AI COE. If neither a decentralized approach nor a centralized approach to a COE worked, the organization’s leaders wondered what would. The answer? Both. To move from orchestrator to problem solver, the COE needed to adopt a federated model, meaning independent AI teams could work to address function-specific challenges while adhering to the standards that the central AI COE set. In this third evolution of the retailer’s AI COE, its federated teams problem-solved in their respective functions. Meanwhile, the central AI COE continued to set coding standards, managed cloud capacity allocation, and standardized AI-related hiring across groups, minimizing duplicative work and making the organization more efficient.
3. Incorporate gen AI into the full product development life cycle
In a typical consumer organization, engineers spend most of their time on tactical work, such as documentation or bug fixes. This can limit an engineer’s availability for the kind of strategic thinking that leads to improved consumer satisfaction. That’s where gen AI comes in.
By incorporating gen AI across the product development life cycle (PDLC), from ideation to deployment, engineers can complete tasks faster, freeing up time for higher-order problem-solving such as improving product design and overseeing AI tool output. Experiments have shown that gen AI can nearly halve the time needed to document and write code.
To help get the maximum value from gen AI tools, however, the organization should establish a baseline for engineering teams’ productivity. The CTO (or the head of software engineering) should measure how quickly engineers deliver outputs and identify the roadblocks that slow them down. Leaders should also track each engineer’s contribution to the overall team, as well as the time engineers spend on inner-loop and outer-loop activities. These baseline metrics can help leaders quantify the value of investing in operating-model interventions. From there, engineering leaders can set KPI targets for those tasks on which developers spend significant time and determine the best gen AI tools to boost developer productivity.
Mercado Libre, Latin America’s largest e-commerce and fintech company, shortened its PDLC by adopting a gen-AI-powered platform that operated within its internal developer platform. This created a unified system for developers to manage applications. So far, the platform has streamlined the work of Mercado’s 17,000 developers and handles 10 percent of customer service mediation on one of Mercado’s retail websites.
4. Strategically insource and invest in tech talent
Investing in tech tools matters little without the right talent to use them. The savviest consumer CTOs already recognize this but may still be making a critical error: Rather than building their tech talent in-house, they outsource the tech roles that might otherwise give them a competitive edge.
Our research indicates that the share of outsourced tech talent is 40 to 50 percent in retail and 50 to 60 percent in consumer packaged goods. While consumer players may find this outsourcing useful in the short term, there are potential longer-term consequences for outsourcing key roles. Outsourcing may be temporarily cost-effective, but it also drains in-house expertise in critical functions such as product management, user interface or user experience design, and data science. These functions are responsible for the seamless, personalized experiences that consumers expect from retailers, and outsourcing them can diminish institutional knowledge over time, limit the skill development of employees, and dull a consumer player’s competitive edge.
Conversely, building in-house tech talent keeps intellectual property under enterprise control, which can lead to enhanced performance thanks to fewer operational interruptions (for example, from constant talent backfills). Retail companies that strategically insource tech talent can see cost reductions of up to 50 percent, since hiring contractors is typically more expensive on a per-head basis compared with hiring fewer, more capable in-house experts. What’s more, the boost in tech employee performance can also help retailers enjoy a 50 percent increase in the speed of tech releases each year.
In addition to sourcing talent in-house, consumer CTOs should consider the latest advances in gen-AI-driven software development and revisit their vendor strategy to determine where it still makes sense to outsource, such as with commoditized skills (like app maintenance) or skills that require specialized knowledge (like high-end design).
Competition for tech talent between consumer businesses and tech players may be fierce, but consumer organizations can still win top talent by investing in purpose-driven hiring, or articulating the organization’s mission and enabling employees to see how they contribute to something greater. Emphasizing partnerships with leading tech players could also attract tech talent. For example, Patagonia, which regularly promotes its commitment to sustainability, partnered with Google Cloud to use data analytics and ML tools to improve the retailer’s supply chain and demand forecasting.
CTOs could also engage in “acqui-hiring” (or acquiring a company for its talent), build their own talent ecosystems (by creating a community of developers across consumer companies to exchange knowledge), or work with talent search firms to close internal capability gaps.
Hiring is only one piece of the talent puzzle. Upskilling employees can build institutional knowledge while helping to retain high performers. Our research shows that retention improves when organizations offer their employees state-of-the-art tools, flexible schedules, time to pursue passion projects, and attractive career paths jointly developed by tech teams and HR.
Mars is taking a multipronged approach to solving its tech talent puzzle. The company is investing $1 billion over the next three years in its pet care division to hire tech workers (including digital product developers, data engineers, and data architects), upskill existing employees, and bring in third-party vendors. Its goal is to double digital sales by 2030.
5. Build platform architecture to support business goals
According to a McKinsey analysis, 20 percent of tech platforms drive 80 percent of the value an organization can realize. Because of this, consumer tech leaders should identify the value at stake for each platform, rather than the cost or scope, to determine which platform to implement first. For example, if a consumer business knew it wanted to improve its customer segmentation, it could prioritize evolving its customer data platform (which would also be prudent, since the customer data platform is one of the biggest value drivers in the greater tech system). Tech teams can build platforms like software companies do—that is, with an eye toward commercializing them. Retail media networks are one example of how consumer players have commercialized new tech products.
One global brand’s digital systems struggled to handle online traffic during product launches, leading to site crashes and lost sales. To overhaul its digital operations, the company adopted a microservices, API-first, cloud-native, and headless architecture to support scalability, faster innovation, and personalization. By using cloud-native solutions, the brand could scale its resources to meet demand, resulting in fewer site crashes, while also optimizing spend by using only the cloud services it needed.
6. Use AI to modernize the tech stack
Many consumer players are bogged down by outdated technology systems. Even retailers that have updated their systems typically have more than 20 percent of their tech assets in “tech debt,” which refers to the “tax” a company pays to address existing tech issues.
Consumer players often postpone dealing with their tech debt because they say it’s too expensive (often hundreds of millions of dollars), it takes too long to do so (five to seven years), it’s too disruptive, or the return on investment is unclear. Some convince themselves that their current systems work well enough as they are.
Thanks to AI, many of these excuses no longer hold up. Today, gen AI can help update code (including the quality of the code and the conversion to easier-to-use programming languages), modernize tech architecture (to boost modularity and support multiple AI agents, for example), remediate or migrate applications to run on cost-effective cloud environments, and provide input and insights into the overall modernization strategy. Think of legacy tech systems like books in Latin; gen AI tools can translate these archaic texts into modern English quickly and cost-effectively.
Using gen AI to modernize an organization’s tech can lead to 40 to 50 percent faster timelines for tech debt remediation and up to 40 percent reduction in costs, while also improving the quality of the outputs.
7. Treat data like a product
Consumer organizations recognize the power of their data, but many struggle to harness its full potential. When it comes to data synthesis, one of two things is typically true: Either an organization has no overarching strategy for data and thus duplicates efforts, or an organization has a central team that manages data—but this team is often left out of conversations about the business’s goals. In either case, the company is leaving valuable unstructured data (such as online customer reviews and pictures in social media uploads) unused. Data architecture upgrades that would support high-value gen AI use cases get continually postponed.
CTOs should instead think of data like a product. Product development requires dedicated management and funding, performance tracking, and quality assurance. So does data. Ideally, consumer organizations would deliver high-quality, accessible data to employees across functions. Our research shows that companies that treat data like a product can significantly reduce the time it takes to implement new use cases for their data, as well as the total cost of ownership. For a consumer player, high-value data such as customer ZIP codes should be managed with data infrastructure and APIs by cross-functional teams, which should include clear data product owners and have joint tech-business responsibility for outcomes.
Nestlé Purina operated several digital properties that required customers to enter the same information multiple times. The pet food maker built a solution that allows consumers to create a single profile that works across Purina’s portfolio. This allowed Purina to deliver personalized recommendations to pet owners based on their pets’ food preferences and age, while improving the user experience. The new system covers 90 percent of Purina’s consumer data collection.
8. Recommit to investments in digital trust
In the mid-2010s, after e-commerce became more prevalent and several high-profile data breaches struck major retailers, cybersecurity was top of mind for consumer players. But as much focus as there has been to embrace new technologies including public LLMs, CTOs must also recommit themselves to building digital trust.
Given the breadth of vulnerabilities in an organization, it cannot be the chief information security officer’s responsibility alone to manage cybersecurity. There are servers, networks, and PCs to consider, as well as cloud devices, operational technology, Internet of Things (IoT) end points, and AI tools to protect. At best-practice companies, the CTO creates—and helps fill—a number of crucial cybersecurity positions, such as data governance manager.
When it comes to data, CTOs should consider whether they have a clear view of their organizations’ most critical systems and data types and what the security challenges are at each stage of the data life cycle.
Sometimes, the biggest threats to digital trust come from inside the organization. Consumer companies that have not updated their AI policies risk declining organizational performance, poor customer perception, and costly regulatory penalties. To combat these risks, CTOs can establish guardrails (for example, deploying fairness tool kits to spot and remove biases from AI models, or masking personally identifiable information before inputting into an LLM). Some consumer players have taken a proactive approach to cybersecurity by adopting advanced threat detection and response systems, improving employee data privacy training, and collaborating with cybersecurity firms to conduct vulnerability assessments.
While these cybersecurity tenets may seem straightforward, consumer CTOs should not treat them as boxes to tick. Instead, there should be regular risk assessments to catch costly digital vulnerabilities before they manifest into real threats.
Consumer CTOs have several high-priority objectives in 2025. They must modernize their technology landscape and reduce tech debt. They must bring gen AI into the entire organization—not just certain teams—to address the entire product development life cycle. They must also cultivate a culture that bridges skills gaps and connects technology with business ownership. It’s a daunting to-do list, but one that, if completed, can separate them from the rest of the pack.
