Issue 16 - Beyond Acquisition

Brought to you by Redo

We tend to talk a lot about ad channels, creative, and other top-of-the-funnel activities. A lot of virtual ink is spilled around acquisition in the DTC space. But it’s only one phase of the buyer's journey, one piece of the puzzle. 

Optimizing user experience becomes more and more vital as you scale. Providing a great CX can be the difference between just “getting a buyer” and “nurturing lifetime value”. 

High-end customer service is a force multiplier for a consumer brand - one that can aid in all aspects of the business - from achieving that first conversion to turning first-time customers into long-time fans. Or even raving advocates. 

For younger DTC brands, great CX is often built on sweat equity and brute force, like the founder personally responding to every inbound call and email. As a company grows, that naturally becomes impossible. High-end customer service at scale is built from systemized conflict resolution, deep user insights, and efficient, tech-enabled processes.

Not too long ago, that meant some mix of human agents, ticketing systems, IVR phone platforms (“Press 1 to speak to an agent”) and/or call centers (“Please hold. We are experiencing a higher than normal volume of calls…”). Clunky, siloed, painful, expensive…probably some mix of all four. 

 The good news is that we’re entering the age of CX evolution. New technology is promising better, faster, cheaper, and smoother customer experience.

 In this week’s edition:

1. AI Agents in the ChatGPT era

2. The value of Zero Party data

3. Optimizing the post-purchase experience

AI of Yesteryear

Remember War Games? 

It’s about a young computer whiz who comes across an AI agent that understands him and communicates intelligently. Of course, tt then goes rogue and nearly causes a nuclear war. 

The movie came out in 1983, about a decade after the first artificial intelligence that could fool people into thinking they were speaking to a human, ELIZA. 

Far from what we’re used to today, ELIZA had pre-programmed responses to keywords designed to make the conversation progress naturally. For example, the word “mother” in your input would trigger a response like: “Tell me more about your family”. 

A number of “AI’s” came out in the following decades. However, machine learning needs massive amounts of computing power and data to reach useful levels of “intelligence”. Chatbots in customer service, which began with IBM’s Watson in 2010, were more of the ELIZA type with programmed responses. 

Until ChatGPT.

AI Agents in the ChatGPT era

ChatGPT can handle a lot of complexity. But did it immediately replace everyone’s customer service departments? 

No. 

Just in January, this guy had a hilariously bad experience with DPD’s AI chatbot.

His experience was a result of the chatbot not being integrated with DPD’s orders database. 

That's probably because it’s complex and resource-intensive to integrate AI agents with large databases. Here are a few other places even new AI agents have typically struggled:

• Context Retention

  Although AI has improved in maintaining context within a conversation, it can still struggle with long or multi-threaded interactions that require an understanding of broader context or past interactions.

• Complex requirements

  Queries that require deep understanding of context, history, and sometimes creative problem-solving can go beyond the capabilities of AI.

• Emotional cues

  When customers display emotions, detecting and interpreting these to generate a response that acknowledges their feelings and addresses them can be difficult for humans, let alone AI. 

• Cost-efficiency

  Deploying enterprise-level AI requires significant initial investment in technology, training, and integration.

Of course, the technology is improving fast… 

Klarna’s OpenAI chatbot made the news recently because of its crazy performance. Here are a few numbers :

• It is doing the equivalent work of 700 full-time agents

• It outcompetes human agents in terms of customer satisfaction score

• It is more accurate in errand resolution, leading to a 25% drop in repeat inquiries

• Customer resolution occurs in less than 2 mins compared to 11 mins previously

• It’s available in 23 markets, 24/7, and communicates in more than 35 languages

• It’s estimated to drive a $40 million USD (!!) in profit improvement to Klarna in 2024

They use the faster-better-stronger ChatGPT Enterprise solution to overcome some of AI’s customer service limitations. 

Now, we have GPT-4o. 

It’s quicker, understands images, voice, and text and can interpret emotions better. Soon customers may be able to speak in natural language to an AI agent or point their phone at a product and ask for a warranty resolution or troubleshooting. 

It seems AI customer service is entering the “actually useful” phase.  

Why You Want It

Because AI agents are scalable as hell. 

• Available 24/7, 365

• Multi-modal

• Multi-lingual

• Continuously improving

Modern chatbots can be human-like in their responses, but they are decidedly inhuman in many advantageous ways. 

You can rapidly train them on your brand’s entire knowledge base (kind of like Neo learning Kung Fu). They don’t take vacations or lunch breaks. They don’t need benefits or health insurance. 

Takeaway: Improving the pace, accessibility, availability, and accuracy of your customer service at scale (and for less) is the promised land for DTC operators/consumer brands. But it’s not the only path to improving CX. 

The Importance of Zero-Party Data

AI agents can now process more data. If enough customer data is available, this means a better understanding of their history, preferences, and needs. With that, you can engineer more efficient marketing, better conversion rates, and more loyalty. 

Enter: zero-party data. 

Zero-party data is “Data that a customer intentionally and proactively shares with a brand”. Seems invasive, but according to PwC, 63% of customers are happy for brands to use their information if it means a better experience. 

So, how can you get valuable ZPD? 

• Quizzes: Ask your users who they are, what they are looking for, and what pain point they’re hoping to solve. Then match them with the most relevant product or experience.

• Preference Centers: Allow customers to set their preferences for communication and product recommendations.

• Pre-and-post-purchase surveys: A goldmine of qualitative information if you ask the right questions. What brought you here? Where did you hear about us? What made you buy? Why did you almost NOT buy? What other products are you interested in?

Bloomreach reports that customers often expect some sort of value exchange when providing ZPD, so think about offering incentives like discounts, gifts, and contest entries in order to supercharge your efforts. 

You can custom-build surveys into your user buy path, but tools like Knocommerce and Fairing now enable DTC marketers to easily collect, collate, and deploy powerful user insights. And remember, we’re facing the end of the third-party cookie, making it more important than ever to proactively collect user data than ever before.  

Takeaway: Understand where your customers are coming from, what they want, and what is causing friction, and then deploy those learnings to improve your CX. Combine with well-tuned AI agents and watch your NPS soar.

Don’t Ignore Post Purchase

Getting that first purchase isn’t the end of your relationship with a customer. It’s the beginning.

Brands that stop caring once the credit card has been charged might save themselves some effort in the short term but risk damaging their reputation, not to mention future profits.

We get it. Customer service, once the parcel has been delivered, can be rife with conflict and feels like a cost center. But a great experience around returns, warranties, lost packages, and product troubleshooting can improve your bottom line over time. 

Here’s why:

• You lower the chances of a return via education and troubleshooting

• You increase the chances of turning a customer into an advocate by smoothly handling complaints and problems

• You increase trust and the chances of a second purchase by resolving issues with speed and ease

• More user touchpoints gives you additional opportunities to promote the brand or upsell/cross-sell. 

As CAC’s continues to rise, it’s vital to provide the best customer experience possible. 

Takeaway: Great CX increases conversion, improves personalization, and expands lifetime value. Leverage the AI boom to scale your CS, survey your users to better understand who they are, and don’t forget to extend your efforts beyond the first purchase. 

Quick hits

• 5 examples of what GPT-4o can do

• Amazon strikes a deal with Meta. You can now advertise for and purchase Amazon goods directly through Instagram.

• Meanwhile, Shopify strikes a deal with YouTube to enable in-video affiliate shopping.

• Heads up! Sophisticated scammers are targeting ad agencies.

• A framework for surviving the ups and downs of startup life.