A marketers explanation of how ChatGPT works.

A marketers explanation of how ChatGPT works.

Mar 13, 2023 | Innovation, Marketing

ChatGPT has blasted into our consciousness over the last 2 months. It has created an equal measure of excitement as people see the opportunities for leveraging their capabilities, and dismay at the problems they see being created.

Both are right, but if we are to make judgements about which side of that fence we choose to sit, it makes sense to understand a little bit about how it works.

These AI tools work on letters, and groups of letters, which then make up words, and the probability of one letter following another, and then another, and then one word following another, and another.

There are about 40,000 commonly used words in English, and billions of words published. From this database computation can give you the probability of a letter following another, eg. The probability of a U following a Q is very high, the probability of a V Following an L is low. This probability logic is extended to groups of 3, 4, 5 letters, one calculation of probability at a time. The outcomes of those cascading probability calculations transforms letters into groups that make up words based on the text used to ‘train’ the software.

Many words have multiple meanings, depending on the context in which it is used, homonyms. Sometimes the spelling is different, but they sound exactly the same. We understand what is meant by the context in which the word appears. For example: if I said, ‘I am on leave’ everyone knows I am on holiday. By contrast if I said, ‘I am going to leave’, it means I am about to depart whatever event we were at. I might also leave something for you at the door.

The word ‘leave’ is spelt and pronounced exactly the same way every time, it is the context in which it is used that makes the difference.

The juxtaposition of words also makes a difference to our understanding. If you remember your primary school grammar, it is all about the position of the subject and the verb.

If I was to say: ‘I am going to leave the party‘ the subject, object, and the verb are in the correct position in English for easy understanding. If I was to say ‘the party I am going to leave‘, most would understand, but would be expecting me to say more, despite the words being identical, it is just the position that changed.

Linguists have studied these relationships for years. Their mantra is: You will understand a word by the company it keeps.

If you take this to its logical extreme, the position of every word in a body of text has an impact on the understanding of every other word, and group of words in the same body.

If the surrounding text to my sentence is about going to a friend’s place for a drink, that will lead to a probability that the ‘party’ has to do with a social event. On the other hand, if the surrounding words were about politics, the phrase ‘I am leaving the party’ takes on a completely different meaning. All these considerations are taken into account by the magic of the probability of me leaving the party when the words friends and drinks are in the surrounding copy. Should those surrounding words be government, and policy, it is more likely the party I am leaving would be a political one.

The operating system of Open AI, and others, have scraped the web for all text published, and stuck it into what amounts to a huge multidimensional spreadsheet. The machine calculates the probability of any one letter appearing after another, then any word appearing next to another based on the occurrences of those letters and words and groups of letters and words in the scraped text. It does this over and over again, spreading the web of probabilities of words and groups of words appearing together, in a particular order, wider and wider, one word at a time, across the body of copy.

This process is extraordinarily computationally intensive. It is hugely expensive to build and program machines that can do these enormous sets of calculations on this amount of text.

If you give such programs a general brief, the best it can do is return a general response. The more detailed you can make the brief, the more explicit the context, the better the machine will be able to use probability to find that combination of words that best matches your requirements, then spit out a response to you.

As a marketer, you understand that when giving a creative brief to an ad agency, the more detail you can give the creatives, the more relevant will be the creative responses. A general brief will give you lots of ordinary creative responses. By contrast, a detailed brief that clearly articulates the target market, product benefits, and the value to be derived from the products use, will generate better creative responses.

ChatGPT is no different, so for good results, give it a good brief.

What makes this so powerful for those who are expert in their domains, is that they will be able to give better briefs, and so have returned better results, which will then be the basis of their creative thinking. This offers the opportunity to improve on the best that has been done to date. For those who are not as expert, their briefs will not be as good, the context in which the machine defines probabilities will be wider, so the output more general, generic, average, and average these days increasingly simply does not cut it.

I hope that helps.

For a more detailed and technical explanation of how ChatGPT works written by an expert, go to the fifth PS at the end of this blog post published when I first stumbled across ChatGPT in December last year.

Header Credit: Dall-E. The brief was ‘ChatGPT algorithms working hard to compute copy in a surreal setting’

5 Myths of referral marketing busted.

5 Myths of referral marketing busted.

Feb 22, 2023 | Customers, Marketing

 

 

Few would disagree that the very best way to find a new customer, to build a business, is to have existing happy customers refer you to their networks.

Even anonymous referrals are better than nothing. How often have you looked up a service provider on social media, and looked at the ratings? Recognising they may be from friends, fools, and their mother, they are still a guide.

Happy customers will refer automatically.

Sadly this is not the case in a proactive sense. Happy customers may give you a wrap when they happen to be talking about whatever problem you solved for them with their friends and colleagues. That is not the same as proactively being an advocate for your product. You have to ask them to refer you, and the manner of asking is crucial.

Customers do not like referring.

In my experience, happy customers do like referring you, but as noted, they have to be asked. The psychological drive of reciprocity comes in here. When you have met and hopefully exceeded the expectations of a customer, they will feel obliged to at least be nice to you. Asking for a referral is a very easy way for them to be nice.

It is OK to pay for a referral.

No, it is not. Paying for a referral is almost an insult. Most people do not like to benefit personally from a referral where there is a friend or acquaintance involved, as it is their credibility at stake.

Potential customers do not believe in referred products.

Yes, they do. When someone who is trusted delivers a referral, that referral takes on an element of the trust that is in the relationship. Both parties know that trust will be damaged if a referral does not ‘pan out’ as promised, so they are careful. This is entirely different to the so called ‘influencer marketing’ that infests digital platforms. These influencers are no different to the talking heads we used to see all the time in ads in earlier times.

Assuming a referral will lead to a sale.

Many things must be aligned for a sale to eventuate, all a referral does is give you a credible foot in the door, the right to have that first conversation in the sales process. You still need to do the hard yards. You still need the sales process.

In a world where the first and must win commercial battle is for the attention of your potential customer, the presence of a credible referral is like getting a 20 metre start in a 100metre race.

 

 

Same challenge, two strategically opposite responses.

Same challenge, two strategically opposite responses.

Feb 3, 2023 | Branding, Demand chains, Marketing, Small business, Strategy

 

Woolworths last week announced they would close 250 of their current 300 in store butcher shops. Clearly, centralisation and opacity of the supply chain that serves customers via Woolworths is geared to the lowest common denominator, price.

At the other end of the scale is Wolki farm in Albury. This is an integrated farm to retail supply chain that innovates at every point. Rather than just trying to do  the same job as always for a lesser cost, they re-engineered the whole chain. From their website: ‘We are the connector between the conscientious consumer and quality produce’

Their 24/7 retail outlet in Albury is just the end of the chain, but full of innovation. I do not normally inhabit TikTok, but this video of owner Jake Wolki’s view of the future was referred to me by a (younger) friend, who knows my views about agricultural supply chains.

The challenge both retailers are setting out to address is the core challenge of marketing: how to create and communicate value that motivates customers to a transaction facilitating longer term engagement.

Woolworths (and Coles, Aldi, et al) do it by price and convenience. They might mumble about quality, but it is at best a second order priority. As long as it is edible, legal, and delivers the category target margin, it is OK. By absolute contrast, Wolki’s (I do not know them at all, had not heard of them until last week) are clearly focussed on quality, product provenance, and integrity. The price they charge for their produce will reflect all that, but no consumer who is looking for the cheapest cut of meat is likely to find it at Wolki’s.  What they do get in detail is supply chain transparency that delivers the provenance and guarantee of quality of the product they are about to buy.

That may interest only a small proportion of the market, but that proportion is significantly larger than it was just a couple of years ago, and will continue to compound.

It seems to me that Woolies are repeating the mistake they made with Thomas Dux 6 years ago. They are ignoring the messages being sent by consumers from the ‘edges’ of their customer base that ‘Mass’ was not acceptable. More probably, they are choosing to ignore those consumers in favour of low cost supply chain control, and reluctance to rock the competitive ship by innovation. Perhaps they will prove me wrong, and use the remaining few in store butchers to experiment?

Photo credit: Wolki Farm from the website 

 

How will AI impact most on marketing?

How will AI impact most on marketing?

Jan 23, 2023 | Change, Communication, Marketing

 

Considering my definition of marketing as being: ‘The identification, development, leveraging and defence of competitive advantage’ it makes sense to consider the impact of AI, as it is happening all around us. Largely unnoticed until the explosive birth of ChatGPT in November last year following the earlier release of Dall-E, the doomsayers are at work.

I am not a data scientist, my limit is writing a formula in Excel no longer than 3 factors, but you do not need to be a data scientist to think about this stuff.

AI learns from itself by iterating with the benefit of ‘digital hindsight’, the outcomes of the previous iterations built in. Think of a radiologist reading scans. In the course of a year they might read a thousand, each time learning from the experience of the previous readings. Over the course of a professional career of 25 years, they might read 25,000, then they retire, and the experience is lost. An AI system can read hundreds of thousands in a week, each building on the previous, looking for patterns, so millions over a couple of years. They can also take data from other sources and blend it into the analysis, and they never retire, so the experience is not lost, it compounds. Importantly however, it compounds based on what has happened, making visible what is already in the data. We have yet to build an algorithm that can be creative.

The ingredients necessary are just 4:

  • Input data,
  • Computing power,
  • Quantitative understanding of human behaviour (still evolving) and,
  • An AI system.

Successful Marketing uses all four, although to date in vastly different ways and to differing degrees. It requires an intimate understanding of customer behaviour and how your  behaviour and that of the customers  impacts others in the supply chain. This is almost ground zero for marketing success.

The combination of the recently released ChatGPT and its stablemate from OpenAI Dall-E will do for content creation in its broadest sense, what the digital camera did for photography. Suddenly everyone became a ‘photographer’, so who needed professionals? Slowly, the gap between even good amateurs and the professionals became clearer, the value added by the real pros, as distinct from the others became more obvious, and presented the clear choices that needed to be made.  A similar process will evolve with written and visual content. It has become very easy to produce stuff that will pass muster as OK, but is that good enough in a homogeneous world?

The combination of these tools and a professional will reduce the time taken to produce great work, so the costs will go down, and the quality will not suffer, but be enhanced. A great outcome for the few true professionals.

The downside will be felt by those who claim expertise, but do not genuinely have it. Their output of regurgitated marketing strategies, tactics and collateral material will resemble the thousands of templates already available, and be of little genuine competitive use.

 

Header cartoon credit: Tom Gauld in new Scientist

 

 

 

The ‘Frame problem’ and marketing

The ‘Frame problem’ and marketing

Jan 16, 2023 | Marketing, Strategy

 

At the intersection of the science of the brain and Artificial Intelligence, is something called ‘The Frame Problem’

This is a term used to describe the way we, subconsciously, sort the relevant from the irrelevant in any context, or ‘frame’.

It locates the inflection point between artificial intelligence, getting smarter by the day, and the sentient intelligence we humans can bring to bear without conscious effort.

Often, we just call it common sense.

For example, if we saw a 3-year-old child we did not know about to jump into a swimming pool, we would automatically try and stop them. By contrast, if we also saw the kids mother waiting a few feet away to catch them, we are unlikely to even register the fact that they are about to jump into the pool.

The resulting ‘frame’ which drives our response is different, although the scene our eyes ‘see’ is identical. It is the interpretation our subconscious makes that is entirely different. That difference is how our brains interpret the factual scene our eyes register on our retina.

Applying the ‘Frame’ to largely qualitative contexts when outcomes are variable, and derived from a host of drivers, frees up cognitive capacity to do other, more important things. In differing contexts or ‘frames’ the variables stimulate differing courses of action, as the value of experience and domain knowledge comes in.

You cannot learn this stuff from a book, as no book can adequately predict which set of variables will show up at any given time in differing contexts. That variability will have a profound influence on the resulting action we take.

For a marketer, understanding the ‘frame’ of their target customer or market will enable you to tweak the drivers that will lead to a desirable outcome. Equally, it will enable discrimination between drivers so that investment is not made in combinations of drivers and situations that will not suit the marketing objective.

The key question to ask yourself is: What did we miss?

 

 

 

A marketers new year resolution.

A marketers new year resolution.

Jan 1, 2023 | Change, Marketing

January 1st is the day we verbalise our introspection.

Usually it is called a resolution, but the irony is resolutions are things we do, and new year resolutions are usually things we would like to do, but in our hearts, know we won’t

Anyway, for a marketer, or indeed any manager, a sensible 2023 resolution will be in three parts:

What should I stop doing?

What should I start doing?

What should I do more of?

Implement that simple resolution set, and 2023 will inevitably be better than 2022, although 2022 was a pretty low bar for most of us.

Have a great 2023.