When A.I. met digital strategy

Two of my favourite areas of interest have been flirting with each other for a while now, and with the arrival of Chat GTP3, things are heating up.

I don’t think its possible to be in any area of knowledge work and not at least cast a fleeting glance over your shoulder at the advances being made in AI. Optimistically with one eye, to see how it can be used as a tool to advance the effectiveness and efficiency of your work, and pessimistically with the other to see how your own role might be effectively and efficiently made redundant.

Hence, I’ve been noodling on the the possibilities of the collision of AI with the customer success model for a while now. But it was the debate at a recent conference which convinced me that the point of impact is much closer than we think.?

In a classic illustration of the Innovator’s Dilemma,?I was with a group of customer success leaders from different SaaS businesses. We were discussing the oft debated balance of ’tech touch’ vs ‘human touch’; how much of my customer success function (for the non SaaS reader, this is effectively account management with a light layering of strategic consultancy) should be done by brute automated CRM software and how much of it requires the delicate empathetic efforts of a customer success manager (CSM) delivered in dulcet human tones?

As an aside - one of the more significant byproducts of Chat GTP's release is the abundance of Innovator’s Dilemma examples; this one is causing more angst than a lightbulb at a candlemakers convention.

During a breakout session, the question was posed; surely the logical conclusion of applying even *some* ‘tech touch’ to your lowest value customer segment is that at some point you’ll be applying *all* tech touch to every, including your highest value, customer segments? And in this endgame scenario will the role of the Customer Success Manager will simply be to take the customer out to lunch??

The push back was strong.?

It was delivered with the sort of patient, but effectively dismissive, sigh that you imagine the horse breeders of Baden-Württemberg in the 1880s might have given when being informed that Carl Benz’s latest novelty plaything of the rich would one day virtually eradicate their industry and reshape modern cities completely.?

In the interest of avoiding a fist fight we moved on to other diverting topics, but the thought stayed with me; as surely as Napster begat Spotify and the decimation of the record labels, so automated emails will spawn the AI management of customer accounts.?

But not the lunches. The lunches will will always be done by humans. As I believe they still are in the music industry.

With the arrival of Open AI’s Chat GTP3 in November, it is clear that both the point and force of impact are closer and greater than we could have imagined.

OpenAI is a research laboratory dedicated to creating artificial general intelligence (AGI) that can benefit humanity. Their mission is to ensure that AGI is developed safely and responsibly and that its benefits are as widely and evenly distributed as possible. They are one of the leading organizations in the field of AI research, and its work has the potential to revolutionize many areas of society, from healthcare and education to transportation and energy.

This will also revolutionise the way content is created, distributed, and consumed. AI could help content creators create better and more engaging content, dramatically reduce the time needed to create content, and automate content creation. AI-generated content could help creators reach more people by helping them efficiently create content tailored to their target audience.

By the way, the last two paragraphs were created by ChatGPT. These last two paragraphs are also lifted directly from a blog post by Mark Schaefer, ?so who owns them? According to Bern Elliot, analyst at Gartner, “it is still unclear”.

Ownership is just one of many open questions prompted by ChatGTP3. The ones that fascinate me are the questions around impact, especially on account management and customer success in SaaS. And from my readings, it will be significant.

The earlier release of OpenAI’s GTP-2 in Feb 2019 had already excited interest in its potential to generate text, provide translations and create summaries of longer texts. Specifically it’s ability to accurately generate paragraphs of intelligible and relevant natural language text in response to a short prompt made it naturally suited to chatbots dealing with FAQ type question-and-response interactions.

But it was the release of GTP-3 in June 2020 that moved things along significantly. With an increase in size (175 billion parameters compared to just 1.5 billion in GPT-2) GPT-3 has a deeper understanding of language and an ability to generate more human-like text. It has also been trained on a much larger dataset, allowing a more comprehensive understanding of language. It also has a more advanced architecture, enabling it to better capture the relationships between words and to generate text that is more coherent and natural-sounding.

I asked Chat GTP-3, the successor to GTP-3, “how is GTP-3 better than GTP-2 for chat bots?”

GPT-3 (Generative Pre-trained Transformer 3) is a language generation model developed by OpenAI that can be used to generate human-like text. It has a wide range of potential applications, including customer support. Here are a few ways that GPT-3 could be used to provide customer support for a SaaS (Software as a Service) company:

1. Automated responses: GPT-3 could be used to automatically generate responses to common customer inquiries. This could include things like troubleshooting steps for technical issues or providing information about billing and account management.
2. Personalized responses: GPT-3 could be used to generate personalized responses to customers based on their specific needs and questions. For example, if a customer asks about a specific feature of the SaaS product, GPT-3 could generate a response that provides more information about that feature.
3. Escalation support: If a customer's issue requires more in-depth assistance, GPT-3 could be used to generate responses that help to escalate the issue to a human support representative. This could include providing relevant information and details about the issue to help the representative resolve the problem more quickly.

Overall, GPT-3 has the potential to greatly improve the efficiency and effectiveness of customer support for a SaaS company by automating many of the routine tasks involved in responding to customer inquiries.

And then in November, ChatGTP-3 dropped. A chatbot built on top of the GTP-3.5. And this is where the radius of the point of impact starts to widen. A lot.

A way to think about AI and tasks is across the axes of repetition (efficiency) and complexity (effectiveness). Roughly you can map the customer facing functions of a SaaS business across this in this way.?

* The level of repetition of technical support questions depends on the state of your product; the more beta and buggy it is, the more unique technical tickets you’re going to receive.

Until Chat GTP3 I felt comfortable allocating AI to the bottom left hand corner of that plane, while acknowledging that it would gradually move up and to the right. And even then, I was pretty sure that it was only going to eat into the ‘Moderate complexity / Moderate repetition: Customer success: Tactical questions’. But a line in Mark Schaefer’s interview ?with Shelly Palmer caught me off guard: “And I’m even thinking about patterns in terms of my consulting practice. You know, one of the beautiful things about marketing is that there are foundational elements that can be applied to lots of different problems. And that’s what AI is good at.”

Having worked as a digital strategist / planner for many years I know this to be true; there are foundational frameworks and processes that are used to structure the approach to marketing challenges. And at the other end of the process I have worked with companies like Spirable and Ad-Lib that use AI and ML driven automation to produce the almost infinite versions of ads required to feed the insatiable algorithms of the social ad platforms and optimise their performance. But what Shafer has touched upon its that generative AI like GTP-3 could be used to solve, plan and execute marketing campaigns from end-to-end.?

And so much as Zeus castrated Kronos, I quietly typed my own emasculation into ChatGTP-3 with these questions;

“How can I codify the elements of marketing strategy planning?”

“What are the data sources I'd need for each of these steps?”

These questions took total of 45 seconds to write and receive responses. So no, it’s not a business pan for an AI marketing strategy platform. But it does give you a sense of what the end-game looks like;

Finally I asked “How could AI bring all of these programs together?”

ChatGTP-3 replied “AI can bring various software programs and tools together by providing a unified platform for data analysis and decision making. Some of the ways AI can be used to integrate marketing strategy planning software include:

1. Predictive analytics
2. Automation
3. Optimization
4. Natural Language Processing (NLP)
5. Machine Learning

By bringing together data from multiple software programs and tools and using AI to analyze and make decisions based on that data, companies can have a more comprehensive and accurate understanding of their customers and the effectiveness of their marketing strategies.”:

Look, It’s hardly Stanley Pollit, (the father of account planning or advertising strategy) but the direction of travel is clear and consistent. Knowledge work that depends on the identification and application of ideas that are based on patterns are going to be replaced by AI. And this includes my area of marketing strategy.

To paraphrase Lord Acton and with apologies to Clay Christensen; “Technology tends to disrupt; disruptive technologies disrupt absolutely”

Now I just need to work out what to do with all of the spare time I am going have on my hands.

Tennis anyone?