Cross-Section of AI and API-Enabled Marketplaces

By Luke Skertich with the help of ChatGPT

Artificial intelligence (AI) has become a buzzword / borderline taboo in almost every conversation (and tweet thread) a VC takes part in. I won’t sit here and preach that AI is the Second Coming. But I do want to share my experience and provide some guidance on best practices.

As with my earlier study on API-enabled marketplaces, I’ve again noticed a follower function with marketplaces incorporating AI plugins a half-step behind B2B SaaS. Architecture certainly plays a role — adoption is easier when you’ve built in a modular fashion. This is why API-enabled marketplaces are set up for success to leverage AI from the outset and compound an advantage in the long run as more and more data is accrued — from personalized user experience to enhanced search to better discovery features and better, personalized recommendations. We’ll delve into how to find the right use case, steps to incorporate a plugin, leverage your data for best results (plus keep it safe) and monetize.

Already exhausted by AI? Maybe I’ll exhaust you by citing Etsy as an example. This is a classic marketplace using innovative AI to understand user intent and provide relevant search results for their customers. For example, Etsy’s AI search algorithm takes into account various factors such as the user’s location, search history, and purchase history to provide personalized search results. The algorithm also incorporates image recognition technology to analyze the images of the products listed on the platform, which helps to identify the products more accurately and efficiently. Additionally, Etsy has also implemented a natural language processing (NLP) system that enables customers to search for products using conversational language. For instance, customers can search for “boho wedding dress” instead of having to use specific keywords like “bohemian” or “wedding gown.” This leads to a distinguished advantage that ages like fine wine. Erhm, gets better as they collect more and more data…

Figure 1. Etsy search for ‘boho wedding dress’

Etsy is not alone. Airbnb, Uber, Instacart, Upwork, TaskRabbit, Thumbtack, and many more marketplaces leverage AI to enhance ranking algorithms, improve user experience, facilitate better communication, unlock the best pricing, optimize routes, and match supply & demand.

Figure 2. Marketplaces applying AI

These companies have successfully implemented AI to enhance their businesses, and other marketplace businesses can follow their example.

So, how do you do it?

1. Identify the specific use cases: This will help you to determine which plugins are most suitable for your business.
2. Research available plugins: Once you have identified the specific use cases, evaluate the available AI plugins that can help you address those use cases. Look for plugins that are easy to integrate with your existing system.
3. Test the plugins: Before incorporating any AI plugin, it’s important to test it thoroughly to ensure that it meets your business requirements. This can involve running pilot projects to evaluate the performance of the plugin in a real-world setting.
4. Integrate the plugins: Once you have identified the AI plugins that are most suitable, it’s time to integrate them into your marketplace platform. Make sure that the integration process is seamless and does not disrupt the existing functionality of your platform.
5. Monitor and optimize the plugins: After the plugins are integrated, monitor their performance and optimize them for better results. This can involve tweaking the parameters and settings of the plugins to improve their accuracy and effectiveness.

Plugins should be chosen based on their ability to solve specific problems and enhance user experience. I believe proprietary data leads to the biggest advantage.

Ok, so, how do you incorporate that into a third party plugin?

1. Understand the plugin’s data requirements: Before integrating a plugin, it is important to understand what data it requires to function effectively. This will help you to determine what data needs to be shared and how it should be structured.
2. Use secure data transfer methods: Proprietary data should always be transferred securely. Encryption is a must.
3. Consider data anonymization: In some cases, it may be possible to anonymize data before sharing it with a plugin. This can help to protect sensitive information and reduce the risk of data breaches.
4. Limit access to the data: Access to proprietary data should be limited to only those who need it to perform their jobs.
5. Monitor data usage: Regular audits and reviews can help to identify any unauthorized or inappropriate use of the data.

It is also essential to ensure that the plugin is secure and that the proprietary data is kept safe. If interested in learning more, check out the appendix!

But, how do you ensure that the data is not used for other customers?

Carefully review the terms and conditions of the AI plugin provider. Some providers may offer options for private or dedicated instances of their plugin, which means that the plugin will only use the data for the customer who purchased the instance. It is important to work closely with the AI plugin provider to understand their data privacy policies and to implement the necessary safeguards to protect your proprietary data!!

But, why do all this work if it doesn’t help you monetize?

Monetizing AI plugins can be done through various methods such as offering premium features, charging for access to data, or even offering AI as a service to other businesses. Etsy’s use of AI has led to increased sales and engagement, which has translated into higher revenue for the company.

Alternatively, you can embrace AI further by leveraging it to optimize prices based on customer behavior and market trends. AI is also optimizing pricing strategies in API-enabled marketplaces through personalized pricing. Personalized pricing offers different prices to different customers based on their purchasing behavior (this is much more complex). Pricing enables businesses to maximize revenue and gain a competitive edge in the marketplace.

In conclusion, AI plugins — shoutout to my favorite AI API in Photoroom — are becoming increasingly popular in marketplace businesses as they can provide a personalized user experience, making selling on your marketplace easier via product presentation, improve search and discovery features, and enhance recommendations.

Then, how do you set yourself up for success?

• Start small: Implementing AI in marketplace platforms can be a complex process. Businesses should start with small, manageable projects and gradually scale up as they gain experience and expertise.
• Focus on data quality: AI algorithms are only as good as the data they are trained on. Businesses should ensure that their data sets are diverse, representative, and of high quality.
• Address bias: Bias is a significant risk when using AI in marketplace platforms. Businesses should ensure that their AI algorithms are trained on diverse and representative data sets and use techniques such as debiasing to reduce the risk of bias.
• Ensure transparency: Businesses should be transparent about how they use AI in their marketplace platforms. They should provide clear explanations of how AI algorithms work and how they are used to make decisions.
• Monitor performance: AI algorithms can be unpredictable. Businesses should monitor the performance of their AI algorithms regularly to ensure that they are delivering the desired outcomes.

By leveraging AI plugins, businesses can gain valuable insights from their data and offer a more personalized user experience that leads to better monetization. It is essential to choose plugins based on the business’s unique needs and goals, ensure that proprietary data is kept safe, and monetize the plugins in a way that aligns with the business’s overall strategy.

If you are a founder looking to leverage AI in their marketplace or are already doing it, please reach out! My email is luke@fjlabs.com. All comments and questions are welcome 🙂

Appendix

On plugin security:

1. Remove personally identifiable information (PII): Before sharing data with an AI plugin, all PII such as names, addresses, phone numbers, and email addresses should be removed. This can be done through techniques such as masking, tokenization, and encryption.
2. Aggregate data: Instead of sharing individual data points, consider aggregating the data to a higher level such as a zip code or industry sector. This can help prevent the identification of individuals while still providing useful data insights.
3. Filter out unnecessary data: (business related data) or data not in the scope of what the requestor should get through the plugin — example from Blobr here.
4. Use differential privacy: Differential privacy is a technique used to protect individual privacy by adding random noise to data. This makes it more difficult for a malicious actor to identify specific individuals in the data set.
5. Check for re-identification risk: Before sharing data with an AI plugin, conduct a risk assessment to determine the risk of re-identification. This can be done by considering the uniqueness of the data, the granularity of the data, and the availability of external data sources that could be used to re-identify individuals.
6. Implement data governance policies: Establish clear policies and procedures for the collection, storage, and sharing of data to ensure that all data is treated in a consistent and secure manner.