Advertising vs. Management Consulting vs. AI

Let's take a quick look at what separates advertising industry from management consulting

(To bastardise Oscar Wilde) Basically every business is three

The business it thinks it is. The business other people think it is. And ultimately the business it really is.

Advertising is the business of designing market perceptions

Advertising deals with shaping what other people think the business is

It isn't the voice of the customer. It is the mind of the customer

And just like Real Estate advertising is in the business of brokering a better position

Position, Position, Position. This is advertising 101

Management Consulting on the other hand deals with what the business really is

Advertising agencies compete on three key qualities

They must be better at insights (positioning), project management and creative messaging than their clients and their competition

Advertising and Management Consultants battle for share of wallet because their competitive edges mirror one another

But ultimately the battle is one of performance vs position

With the right position the business doesn't need to outperform the competition

With the wrong position businesses have to outperform the competition

Put simply any business spending excessively on management consultants is doing so because it has failed at marketing

It is investing in improving its own performance over improving its market position

Maintaining a poorly positioned business is a money pit

The overwhelming majority of businesses are poorly positioned

This is why management consulting is a growth industry and advertising is in decline

The trick that adtech and digital pulled was to conflate performance with position

Convincing the marketing world that performance - rather than being a measure of a brands relative position - improves position

e.g. High CTR = Stronger Position - where in reality a strong position (in the customer's mind) not only delivers a higher CTR it also deliver search and social traffic because share of mind translates into share of search and voice

Today marketing managers choose to lease their position (on the list - be it Google or Social i.e. Virtual memory) rather than own their position (in the customers mind)

in 2000 I identified the value proposition for advertising moving forward was customer intimacy

ie. We decided we were in business of identifying and designing tactical opportunities that will enhance our client's Customer Value Proposition and/or Customer Life-Cycle Profitability

That business was about developing relationship strategies. Both digital and IRL

The future of business was interactive

It was an attempt to merge advertising and management consulting into a single offering

It moved the business away from the 30 sec spot or web site design towards customer journeys and business process re-engineering

In a word the future was digital transformation

The pitch was Mission Critical:Tactical Initiatives delivered in 90 days

It was called Decision Point 90

The proposition looked something like this...

The balance between the Yin Yang was a reflection of the relative business maturity of the client

Management consulting profits from novice level best practices

Advertising profits from clients competing in crowded markets

20 years on the integrated 'digital transformation' offering is the default offering of the consulting factories (eg Accenture)

The risk, or more accurately the weakness, with the DP90 model was it addressed the 4% market share at the tip of advertising and management consulting pyramid

ie Putting in highly skilled teams into high risk/high return ventures within large corporates seeking to role out a new set of best practices designed to capture a new market category

It was a paradox at so many levels

I admit in retrospect we made both massive gains and massive losses in the transition

We shifted from being in the business of improving your market positioning to improving your operational performance

and something fundamental was lost in translation

This vision became a series of massive projects, with restrictions on budgets and teams that were just too hard to deliver... never mind sell without the quick wins we previously made in repositioning our client's brand

We forgot the trick of opening the performance floodgates was repositioning the brand

Without the online business loan calculator the opportunity to pitch the redesign of business banking from the ground up never arrived

Needless to say it's time to get back to basics...

Now I would argue management consulting evolved from 'taking your watch to show you the time' to 'taking your watch to show somebody else the time' sometime back

Thanks to the wonders of digital and the zero cost of distribution they can now 'take your watch and show everybody the time... in real time'

Which is to say their economic function was never about 'invention and evolution' of new ways of doing business but the 'replication and contagion' of existing ways of doing business

The billable hours being of course a measure of the contagion

Within this context the assumption that the profession can become innovative is oxymoronic simply because this is neither its function or collective capability

i.e. its function is to share (& therefore accelerate the adoption of) best practices not create them

The advertising profession lumbers under a similar - dare I say parallel - misconception about its creative function in the business world

Advertising's function isn't to generate new ways of seeing or new ways of doing

Advertising's function is to distill the zeitgeist and then replicate and amplify it

In this advertising and management consulting are mirror activities

They analyse what works - what is hot - and amplify it so it quickly becomes the mainstream - the defacto, the default way of thinking and doing. Of seeing the world

These are not creative industries. They are hypermemetic industries

ie In the digital age they engage in the extreme and unrelenting vomiting of performative memes and xemes to advance their client's interests

The digital revolution has also brought automation to both sectors

Consulting has evolved into Software as a Service and Advertising into AdTech

What was once a high cost, high touch service has become a low cost, just in time, tactical commodity

Today we have Position as a Service (Think: Google & Instagram) and Performance as a Service (Think: Salesforce)

Digital also brought us the rise of the digital marketplaces

and once again what was once a high cost, high touch service has become a low cost, just in time, tactical commodity

Inevitably project management becomes the core competency of the consultancies - no matter what their specialisation

That's why 20 years on the consulting services industry leaders are best understood as branded project management entities

Dominating the 60% of the client pool that is shopping around for best practices in an established, crowded market (ie the Low Hanging Fruit)

The reason being they are trusted to deliver on the promise (of delivering the project on time, in full and on budget) - the fact they rarely do is another story...

Their value driver? Operational Excellence

What we got right with Decision Point 90 was the focus on project management, the building of the talent marketplace and the transitioning of client's towards the automation of the new business processes

What we got wrong was we focused on then pointy end of the business rather than just harvesting the quick wins

While everybody was getting paid just running around the track in circles we tried to take on the pole vault while simultaneously reinventing it as a team sport

and this leads us to the inevitable question... looking forward 20 years what will change?

The easy answer is AI will have triggered a rethink in how creativity is procured (See Advertising & the emergence of the Xeme Factory)

The only question marks will be: Will it trigger a rethink about how Insights and Project Management are procured?

My assumption - assuming it can get Insights as a Service and Project Management as a Service to work - is Ai will come to dominate the low hanging fruit

Suggesting the market leaders invested in the factory model are most at risk from this technology

Meanwhile the hard stuff. The rocket science. It will be left to the rocket scientists

I suspect we'll see an explosion of micro or nano consultancies scattered across very narrow, highly profitable industry verticals

ie individuals or small teams empowered by Ai tasked with solving micro or nano performance and positioning problems

The key measure of AI's impact on management consulting and ultimately SaaS will be its ability to create new metrics

Just how effective will it be at coming up with new metrics and lists to improve business performance and operational capability?

Can AI come up with a better accounting and valuation metrics beyond EBITDA or the PEG Ratio. Can it design a KPI/Bonus structure that aligns management renumeration with shareholder interests?

This is the creative potential of AI

Its ability to redefine how business measures performance and therefore organisational behaviour

SaaS is basically "Spreadsheet as a Service"

The Spreadsheet ingests, digests and then excretes a new metric or list

The typical entrepeneur creates one new metric or list that becomes the basis of their business model

Successful entrepeneurs tend to generate just one metric or list (eg Page/Brin & Google/Pagerank or Zuckerberg & Facebook/Connections) the successful monetization of which allows then to acquire more lists or metrics (eg Youtube and Instagram)

Those that do iterate tend to do just that - iterate (eg Duffield - Peoplesoft and Workday)

AI on the other hand has the potential to create an infinite number of new performance metrics and lists

The game then becomes one of how quickly you can automate the gap between the invention of the new metric through to adoption and ultimately the automated expression of the metric via SaaS

Suggesting, within the context of high margin, premium management consulting, AI as a Service is less about Chatbots and more about Metrics as a Service - or more accurately Benchmarking as a Service

and in this context the new Decision Point 90 looks like this

Basically an emergent cross pollination model powered by AI

But the potential of this model has to evaluated within the context of the maturity of the current AI models being promoted by OpenAi and the other ML/LLM researchers

It may be magical but is it transformative?

ATM all the evidence suggests the creative side is strong but the logical side of the model is weak

More Artificial Imagination than Artificical Intelligence

The question is will this change? and, if so, how long before it does?

and finally does it matter?

Does AI need to outperform in all 4 quadrants?

Arguably not...

But it does raise the question will people believe in the magic if they can't trust where it comes from?

and this leads us back to the robo management consultant experiment we conducted back in 2020

The one where I stumbled across a Tweet describing McKinsey's 'Rule of Three' for crafting persuasive arguments

The rule states: You need to provide the client with 3 reasons to adopt your recommendation. Not 2, Not 4. Just 3

... and this piqued my interest

McKinsey's consulting model generates in excess of $1 Million per employee

Just how hard would it be to create a Robo Management Consultant that replicates McKinsey's advisory model?

Textural analysis of the HBR suggests the question can be resolved if we can reduce the activity of the management consultant into a value equation

For example...

The objective of the consulting assignment is to...

Increase [w] by Reducing [x] and/or Improving [y] by Changing and/or Creating a new [z]

Having developed the value equation we can then randomise the outputs based on the assignment objective

For example...

We can increase

(Choose an Option)

Profit Revenue Growth Market Share Retention ROA ROI Efficiency Cash Flow Visibility

Answer:

Of course this isn't AI. But with a user questionaire (ie Please rate the quality of this solution out of 10) creating the feedback loop it becomes the basis for generating an AI training set

Add a chatbot engine, once it has been trained and hard wired into Google Trends and there you have it... A veritible McKinsey for the masses

However the deeper observation I want to make is this simple random solution generator illustrates how, what passes for strategic thinking in the management consulting world, is in reality, just combinatorial word play targeting the collective pareidolia (The psychological phenomenon of discovering patterns in randomness) of the C-Suite

Suggesting to be transformative in the world of Management Consulting AI simply needs to be magical

ie disrupt the market for generating new metrics and performative ideas

In other words position, position, position

Leaving us with something to think about perhaps... at least until next time

Postscript:

October 2023

Rethinking the Consulting Services Value Chain

‘A consultant is somebody who borrows your watch to tell you the time’

You have probably heard that quip before

And so it is probably the ideal anchor to provide context to a Harvard study that has been doing the rounds in recent weeks

You know the one. You can see the evidence in the chart below.

It provides definitive proof that consultants using AI are more productive than their counterparts

Now I have commented elsewhere that arguably the flaw in this study is the assumption that consultants are hired for their productivity & creativity when in truth consultants are hired to de-risk the situation

If the study provided proof AI can de-risk the situation faster, cheaper or better (ie solve the triple constraint) then one could assume the study provided a valid insight into the utility of the technology... otherwise it just makes the case for AI being (yet) another dimension to ITs proven ability to 'make new work' rather than resolve the underlying business challenge

E.g. does the ability to create a 1000 business cases or powerpoint decks in a fraction of the time really solve the underlying challenge of a company wide inability to innovate or commercialise new products?

But what of the question of watches (think: clocks), the role of consultants and the potential for AI to change the game?

The origins of the management consultant can be found in the first golden age of American innovation. A time of electricity, machines & production lines. They earned their reputation doing time and motion studies: Hence the reference to the watch/clock.

They were the gurus of productivity and efficiency. Master of conquering the triple constraint - How to do it Faster, Cheaper, Better, 24/7.

They were not the watch makers. But they knew how to fix a broken watch.

You could say they are the watch repair men (and women) of the corporate world.

And so to the AI conundrum.

In this context AI needs to learn how to fix a broken watch faster.

It needs to figure out how to get an organisation & it systems running like clockwork.

and this takes us back to the original observation

‘A consultant is somebody who borrows your watch to tell you the time’

You see the Harvard study resolves the question: Can AI help consultants tell the time quicker? The answer being yes.

But in truth any consultant can tell you what the time is. Some of the more experienced may even be able to tell you how many clocks you have. Some of the better ones will tell you which clocks are broken & maybe even get them ticking again. But it takes the truly great consultant to get all your clocks chiming in unison.

and in this context I'm just not sure AI will disrupt consulting by doing things faster, cheaper and better.

If it does disrupt the industry it will make it - at least as we know it today - obsolete by doing things differently

It will become the new watch maker…

But in the meantime I think we’ll see a lot of activity being played out across the value chain as the incumbents jostle for position leveraging AI to grow market share

How will this play out?

Well the illustration below maps the 'value' of the service offering to the type of engagement. The key idea being the industry leaders are in the business of delivering transformation while the rest are in the business of feeding the Zeitgiest (ie Profiting from FOMO - the client wants one because everybody else has one)

The underlying assumption is AI (or at least #GenerativeAI) will disrupt those feeding the Zeitgeist and the focus of this disruption will be at the Factory level as the players in this sector are 'pinched' by Freelancers & Consultancies using AI to automate their workflow

Having said that there is a real risk that AI currently doesn’t have enough of a force multiplier to achieve the desired outcome currently being championed by the AI gurus

As Gearóid Carroll has said elsewhere 'Disruption seems to be an addiction' and within the wider context of the ongoing Silicon Valley narrative clearly this is the case. The deeper question must be however is this addiction enough to change the world of consulting?

To answer that question let’s take a look at Aku Nikkola’s precis of the HBR paper

"Harvard Business School has just delivered the most pivotal working paper to date on the impact of AI on knowledge worker productivity and quality.

In a nutshell, a team of leading social scientists examined how ChatGPT-4 affects the daily work of Boston Consulting Group (BCG) consultants, a substantial 7% of their consulting force (758 consultants).

The results are crystal clear: consultants using ChatGPT-4 outperformed their counterparts in every way. They took on 18 different tasks, carefully chosen to reflect the typical work at an elite consulting firm, and the AI-equipped consultants excelled.

→ They finished 12.2% more tasks on average, completed tasks 25.1% more quickly, and produced results of 40% higher quality compared to those without AI support."

So the productivity gains are Faster/Better

Now let’s put these improvements up against the default benchmarks of the networked economy - email and SMS

When it comes to being faster email and SMS offered a force multiplier of 3000x over the paper incumbent (think: Post Office)

In this breakthrough study #ChatGPT 4 delivers a force multiplier of just 1.4x

Which is to say - at least within the history of the disruptive networked economy - somewhat underwhelming

And we can map this technology again other innovations… e.g. the iPod - the 1000 songs in your pocket vs 20 on your Walkman cassette player - a 50x force multiplier vs just 1.12x for ChatGPT4 in the hands of BCG’s consultants

Then there is the everything stores of Amazon and Alibaba with their force multiplier >10000x over the average department store

and you start to become aware of the gap between the disruptive narrative and the hopeful (or should I say Incremental) narrative. At least when it comes to the potential impact of this new wave of technology on the global networked economy

and this in the end is the lesson in business strategy that ChatGPT and Generative AI teaches us

It's all about the questions we ask to forecast the potential ROI

In this case the question is what is the force multiplier?

Because without being 10x faster or 10x cheaper or 10x better or 10x different what is this technology really bringing to the table?

So let's wrap this up with what I believe is the deeper insight to be found at the end of this journey

Arguably we have found a new innovation metric

The Disruptive Force Multiplier = Speed x Cost Savings x Quality x Differences

Or, DFM = F x C x B x D

So, in the context of consulting services, we discover the ChatGPT 4 DFM equates to this...

DFM = (1.25 x 1.12) x 1 x 1.4 x 1

Or, 1.96x

Suggesting an investment in ChatGPT will make your business almost 2x better off than the competition who doesn't make the effort to embrace GenerativeAI

Which is something to think about I guess...

The Easter Egg

I spent an hour compiling all the recent studies into the impact of GenAI on organisational prductivity

A pattern started to emerge suggesting you could have speed or quality but combining both delivered anemic results

This came as something of a surprise

The data also suggested the technology offered something of a 'Right Brain' bias

This supported my earlier observations about the technology being ideal for Augmented Ideation - so no surprise there

However the data did provide another dimension to the puzzle

One that I had previously canvased but hadn't seen any data to support it

One of the great lines I've read about GenAI is "Why would I waste my time reading something nobody could be bothered writing?"

It is an expression of creative angst - but there is probably enough data begining to emerge to suggest it may well become something of a truism

You see the studies suggest that although GenAI improves the quality of the individual creative it has a negative impact of the overall quality of the group

Which is a nice way of saying everybody tends to generate the same type of stuff using GenAI

With that in mind we now have 4 DFM scores rather than just one to consider

We begin at the individual level

Our DFM score for the creative individual looks like this

1.4 (Quality) x 1.12 (Speed) x 1.25 (Cost) x 2 (Difference) = 3.92x

Not bad... I can see the value in that

Our DFM score for the business process consultant looks like this

0.71 (Quality) x 1.01 (Speed) x 1.25 (Cost) x 1 (Difference) = 1x

and you begin to wonder: Why bother?

Now let's look at the organisational level

Our DFM score for the creative groupthink looks like this

1.4 (Quality) x 1.12 (Speed) x 1.25 (Cost) x 0.59 (Difference) = 1.8x

We have just halved the ROI by introducing the risk of everybody looking the same

Our DFM score for the business process groupthink looks like this

0.71 (Quality) x 1.01 (Speed) x 1.25 (Cost) x 0.59 (Difference) = 0.6x

Run away... because you are introducing significantly more risk to the assignment

Now you may disagree with the science and the maths behind the DFM metric

But It offers a new way of evaluating the potential value of this tech and surely that's enough to start the debate over the wider economic impact of this technology on staff and group productivity?

and so onto our final question...

Just how exceptional are the finding from these early studies into AI productivity improvements when placed within the broader historical context?

Because only then will we be able to quantify the broader potential economic impact of this technology

The data exists to help us take a stab at this question

The sources are
How People Can Create—and Destroy—Value with Generative AI (2023)
The social economy: Unlocking value and productivity through social technologies (2012)
FRED: Nonfarm Business Sector: Labor Productivity (Output per Hour) for All Workers - 1950-2023
THE EFFECT OF C OMPANY SIZE ON THE PRODUCTIVITY IMPACT OF INFORMATION TECHNOLOGY INVESTMENTS (2006)

Here's the mash up from all those sources with the HBR graphic

As you can see the future comes down to an old familiar question: Does AI have network effects? or, something very similar, because without it you are not going to get the productivity boost the introduction of the networked economy brought to the US or the Global Economy

Further Reading -

Advertising & the emergence of the Xeme Factory

Can AI be the new Stock and Schlock of Advertising?

Can StableDiffusion be your new art director?

The Innovation Marketplaces

Boxes and Triangles

Originally Published Summer 2023. What are we talking about today? Follow us on Twitter