Will Your Brand's AI Agents or Someone Else's Serve Your Customers?
This is the third blog in Yali Sassoon's Agentic AI series. Click the links below to access the other blogs:
Brands: whose AI agents do you want engaging with your website and applications, on behalf of your customers? Yours? Or someone else’s?
Imagine a luxury AI agent specializing in skin care. The agent is singularly focused on the health and vitality of your skin.
It is able to draw on deep medical knowledge and access the latest in skincare research and product development. Best of all, it understands you.
It knows your brand preferences, lifestyle, and skin type. It also sees how well different skin products have worked for you so far. Drawing on that knowledge, it informs and guides you to buy the right products, so you always look your best.
Now the luxury AI agent is about to recommend a skincare product. Do you want it to be your brand? Or someone else’s? What say do you want to have in what it is recommending to whom?
The deployment of AI agents will transform the way that customers engage with brands. I use the term "brands" broadly. This includes any brand that faces consumers or businesses, like retailers and FMCG companies.
Brands have the opportunity to re-imagine entire customer journeys and build agentic applications that make these journeys more enjoyable and rewarding, driving customer and commercial value.
Brand-owned agents are uniquely positioned through their deep knowledge of the customer. At the same time, other companies, including OpenAI with Operator, have the opportunity to build customer-facing agents, and have those agents engage with your website and mobile apps on behalf of your customers.
Therefore, brands need to ask: whose agents do they want engaging with their website and applications on behalf of their customers?
- Their own?
- Other companies? (E.g. OpenAI’s? Your competitors?)
- Some combination of the above? (Hybrid approach)
To help you decide, check out the matrix below.
What is at Stake?
There is a huge amount at stake. If you invest in customer-facing AI agents, you stand to gain more loyalty, differentiation, and market share. If you don’t, you risk AI-native middlemen inserting themselves between you and your customers, chipping away at your brand’s relevance.
Option 1: Building Your Own AI Agents
There are four good reasons why a brand should make sure its customers engage with it via its own agent(s) / agentic application(s):
- Owning the end-to-end customer experience,
- Creating a differentiated user experience to drive competitive advantage, e.g. by leveraging proprietary customer knowledge
- Deepening customer relationships, and
- Gaining richer customer data and intelligence.
1. Owning the End-to-End Customer Experience
Customer experience is a big part of a brand promise. If you are a luxury beauty brand, the experience of engaging with your content and buying your products should feel luxurious. If you are a financial news company, your content must reflect your commitment to veracity, timeliness and depth of coverage.
To ensure your brand promise is kept during the customer journey, you have two options. You can control the process yourself. Alternatively, you can generate enough influence over the partner who is in control of that journey.
2. Creating a Differentiated User Experience to Drive Competitive Advantage
Agentic applications provide the opportunity to build differentiated customer experiences that reinforce the brand values. They can do this because they know their customers well (having built proprietary customer intelligence on top of their proprietary customer data) and being experts in their categories. By providing a better agentic application experience, brands can build competitive differentiation and use this differentiation to grow market share and customer loyalty.
3. Deepening Customer Relationships
Customer relationship is a key component of the enterprise value of brands. The birth of the internet allowed many new brands to grow. Organizations learned to build strong, direct relationships with customers online. Before, these connections were often made through retailers
The birth of agentic applications means that those relationships can be developed in ways that were previously unimaginable. Now, brands can talk directly to individual customers at scale, enabling brands to get closer, one-to-one with each customer.
4. Gaining Richer Data and Customer Intelligence
Agentic applications create enormous amounts of invaluable customer (and agent) data. This data is key to better understanding customers over time. It also helps to improve performance in the customer-facing agentic application itself (via a data flywheel e.g. reinforcement learning).
Because owning the agentic application experience enables brands to
- deliver on their brand promise,
- differentiate in user experience, growing market share, and
- continue to own the customer relationship, whilst developing and deepening it, and
- collect more customer data and develop better customer intelligence
I expect most brands to invest heavily in developing their own customer-facing agentic applications. However, there are challenges with this approach.
Brands will have to learn how to build and run customer-facing, workflow-oriented agentic applications, and use those applications to develop and deepen customer relationships.
This is a very significant undertaking - one that many brands will struggle with.
Brands will end up competing with a host of new companies
Brands that develop their own customer-facing agentic experiences will compete with OpenAI’s Operator and other providers of similar services.
Competing in agentic application development with AI-native companies will be tough. These companies, especially those with OpenAI’s resources, have a big advantage. They understand the next generation of agentic technology well. They know how to use it to create great customer experiences. Nonetheless, brands have a number of advantages on their side:
- A better understanding of their customers (and especially what types of experience and service their customers will find compelling)
- Brand loyalty and customer trust
- A better understanding of their space / vertical. (OpenAI’s Operator knows very little about skincare, for example.)
I will explore these advantages in more detail in a forthcoming blog post.
Option 2: Working with 3rd Party Agents
Rather than build their own customer-facing agentic applications, some brands might choose to instead focus on being the best place for agents to engage with their type of product and service, and compete for customer market share through agent market share.
Friend rather than fight AI-native companies
The primary benefit of this approach is that brands do not have to compete in agentic applications. This is a whole new playing field for them. Instead, they can partner with companies that are experts at it, like OpenAI.
But at the risk of disintermediation
This is a risky strategy, however. The brand effectively gives up the customer relationship by allowing itself to be disintermediated. It now no longer can guarantee the brand experience to the end customer. It cannot ensurethat its products and services will be the ones recommended by the agent.
And a lack of control over the presentation of your brand
Can your brand trust an AI to represent it faithfully? One challenges with agentic applications is making sure they act in line with your brand. How can agents that represent many brands do so honestly?
Brands will have to learn how to optimize their website, applications and APIs for agents i.e. Agent Experience Optimization (AXO)
If brands are going to rely on third party agents to engage with the brand on behalf of their own customers, they must learn how to deliver an excellent service to those agents. If they don’t the agents may decide that they can deliver a better service to those customers by going to a competitor brand instead. Brands that follow this strategy must therefore invest in Agent Experience Optimization (AXO).
Option 3: A Hybrid Approach
Given the rewards, risks and complexities of both of the above approaches, it is likely that many brands try to do both in parallel i.e.:
- Develop their own customer-facing agentic applications
- And work to play nice with third party agents
This approach gives brands two opportunities to win market share.They can create their own differentiated proposition. They can also work with third party agents to try and attract as much of the demand from those agents’ customers.
The primary disadvantage of this approach is all the resources required. The brand will need to develop its own customer-facing agentic applications, and its own Agent Optimization Experience program.
Ultimately, however, my view is that brands need to invest in developing customer-facing agentic applications. This will help them to preserve and deepen their customer relationships. It will also prevent themselves being disintermediated by third parties like OpenAI. Though brands should still work with these partners to attract more customers.
Brands already have strong knowledge about their customers.They build these deep relationships over time and these valuable assets have been hard wondeep - they are not going to want to give them up any time soon. As a result, developing customer-facing agentic applications is a strategic imperative for brands.
Option 4: Wait and See (No Immediate Action)
Given how fast the agentic application is evolving, brands could be forgiven for punting a decision. They may want to wait and see how the industry develops before making any investments.
I would argue that this is the worst of all worlds, resulting in the brand falling behind competitors in capabilities around:
- Agentic application development
- Agent Experience Optimization
So What Option Should You Take?
The situation brands find themselves in today is not dissimilar to when travel companies, for example, found themselves contending with the rise of the travel aggregators (Booking.com, Expedia, Kayak). When faced with a choice between working with an intermediary (with access to lots of demand) vs developing their own digital (but not agentic) experience, the vast majority of brands went with a hybrid approach.
In a similar vein, I’d expect the vast majority of brands across industries to go for a hybrid approach to agentic customer experiences: both developing their own and partnering with the likes of OpenAI, who are bound to control varying levels of demand across different industries.
Whilst this is a lot of work, the future of the brand is at stake. The industry is moving fast. Organizations need to start learning both how to build customer-facing agentic applications and optimize their websites and applications for AI agents as soon as possible.
Are you building customer-facing agentic applications?
At Snowplow, we’re developing the next generation of data infrastructure for customer AI. We’re partnering with some of the most forward thinking brands, to enable them to develop agentic applications that better understand and serve their customers and optimize their websites and applications for AI agents.
If you are interested in learning more then get in touch.