Episodes

Harry Davies: The Installed Team

Harry Davies, VP of marketing strategy, investment and effectiveness at Sage, on why brand marketing wins the buyers who are not in market yet, and why AI could let B2B companies install a marketing team rather than hire one.

 ·  The Business of Marketing  · S5 E110  · 48 min

"A lot of businesses still think marketing is coin operated like sales"

A career on the scientific side of marketing, from randomised direct mail tests for central government to measurement roles at Google and Amazon, now leading strategy and effectiveness at Sage. Davies argues that most companies measure only the promotion they can see, that brand exists to win buyers years before they enter the market, and that AI is about to let B2B companies install a marketing team rather than hire one.

Davies started as a fine art graduate who needed to earn a living and fell into marketing, then got obsessed with it after a talk on direct mail testing. He printed two versions of a recruitment letter, one offering a free bottle of water at enrollment, randomly assigned them across 2,000 envelopes, and found the voucher lifted turn up rates. That test and learn instinct led him to a postgraduate in economics and mathematical science, then to central government working on econometrics, including whether marketing could get long-term smokers to quit. From there he moved through Google and Amazon and now leads marketing strategy, investment and effectiveness at Sage, always on the scientific side of marketing evaluation.

In this conversation with host John Horsley, Davies argues that marketing effectiveness has collapsed into media effectiveness, ignoring the rest of the mix where the bigger gains sit. He makes the case for brand advertising as a way to reach software buyers who only come into market every five to eight years, so Sage is first on the list when they do, reframing the debate for sales and finance as building the pipeline of today and the pipeline of tomorrow. He is bullish that AI can decouple revenue growth from headcount growth and even let companies install a full service marketing team rather than hire one, while warning that distinctiveness is still hard and that short CMO tenures wreck the long-term consistency marketing needs. His throughline is simple: read the science, live with your customers, and hold your course.

  • Davies came to marketing sideways, as a fine art graduate who needed an income, then turned it into a science. An early randomised test, two versions of a recruitment letter with one offering a free bottle of water, showed him you could measure almost anything, and that conviction carried him through a postgraduate in economics, into central government econometrics on problems as hard as getting people to stop smoking, and on to Google, Amazon and Sage. His whole career sits on the evidence-based side of marketing evaluation.
  • His central critique is that marketing effectiveness has quietly become advertising effectiveness. As media grew more technical, teams poured their attention into the promotion mix and got very good at it, while product, price and placement, arguably more important, got far less scrutiny. He also warns that where measurement lives matters: an effectiveness team inside marketing tends to trawl for evidence that spend worked, while one reporting through finance or operations gives more honest answers because it made none of the decisions.
  • The brand argument is grounded in buying behaviour. Software buyers only come into market every five to eight years and you cannot control when, so if you do nothing until they arrive you have to outspend competitors to win them. Reaching them at their category entry point, so you are first on the shortlist, is cheaper. Davies translates this for sceptical sales and finance leaders as building the pipeline of today and the pipeline of tomorrow, using their own language of quota, pipeline and recognition rather than brand theory.
  • On AI, Davies has watched tools give sales reps a day a week back and make them so well briefed that a prospect assumed one had worked in his industry. His conclusion is that the prize is effectiveness, not just cost cutting: AI can decouple revenue growth from headcount growth, and for many B2B firms it could build value propositions, define the ideal customer profile and pick channels better than under-trained marketers. He goes as far as saying companies may install a marketing team rather than hire one, while conceding FMCG still needs human creativity to differentiate near-identical products.
  • His closing worry is time. Marketing is not that complicated, you do the right things consistently and the benefits compound, but short CMO tenures and quarterly reporting push people to change everything at once and abandon course before it works. Winning market share is a five to 10 year job, not a next-financial-year job, and doing it needs board trust and evidence rather than the assumption that marketing is coin operated like sales. Read the academic literature, spend an inordinate amount of time with customers, and hold your course.
  1. 01 Marketing effectiveness beyond media
  2. 02 Brand and not-in-market buyers
  3. 03 AI in B2B marketing
  4. 04 The full customer lifecycle
  5. 05 Long-term consistency and CMO tenure

Key Exchanges

05
01 You have moved from a demand generation model to a more balanced brand approach. What prompted that?

If they come into market thinking about us first, it's a lot easier and a lot cheaper to convert them

In B2B software as a service, buyers come into market every five to eight years and we cannot control when. If we do no marketing while they are out of market, then when they arrive we have to outspend competitors to win them. If they come into market thinking of us first, it is easier and cheaper to convert them. So my argument for more brand marketing is to reach those not in market buyers at their category entry point, when they start thinking about finance, with a consistent message about their pain points, so that when they do come in, randomly, we are first on the shortlist.

02 How are you optimising for the large language models and using AI in sales and marketing?

We shouldn't be thinking about how do we cut costs. We should be thinking about how do we drive more effective marketing with these tools

We enabled sales development reps with AI tools, which overnight saved a day a week each on research and turning call transcripts into handover documents. Then I listened to their calls. On one, the prospect asked the rep if she had left catering manufacture to join Sage. She never had, but the prospecting agent had briefed her so well on the industry that he assumed she must have worked in it. That is when I thought we should not be thinking about cutting costs, we should be thinking about driving more effective marketing with these tools, and cutting costs. You can keep quality very high while decoupling revenue growth from headcount growth.

03 Could AI change how B2B marketing teams are built?

You install marketing as opposed to hire a marketing team

Many B2B companies could install marketing rather than hire a team. Give an LLM your product release documents and prospect information, and it will do a better job than unqualified marketers at building value propositions, defining the ideal customer profile and choosing channels, then automate it. In FMCG, where products are near identical, you still need human creativity to differentiate. In much of B2B, especially early stage, you can automate almost all of it, and at a better than average level.

04 Is there a danger of sameism when B2B businesses use AI to craft content?

I could just change the logo on that, and nobody would know

There is, but I am not sure the industry is not already doing that. I watched a brand marketer at a conference say all the right things about distinctive assets and repetition, then show a video, and I thought, I could just change the logo on that and nobody would know. Lots of people have read the Ehrenberg-Bass work and understand distinctiveness, but delivering it is really hard, partly because everyone has an opinion and distinctive assets often turn people off precisely because they do not look like the category, which is the point. AI risks being a bit bland, but it is not replacing some world where every B2B brand is already vividly distinctive.

05 Do a lot of B2B businesses concentrate on acquisition rather than the full lifetime?

Most B2B businesses consider marketing as the acquisition engine. They don't necessarily think of it as the recurring revenue engine

Definitely. I like the Winning by Design bow tie: a funnel from brand awareness to the mutual commit, then retention and expansion on the right, delivering recurring impact. People get that from a sales and customer success angle, but marketing has a role across the whole thing. Keep marketing through the pipeline and you accelerate it and defend against competitors, keep marketing to existing customers and you keep driving value. Most B2B businesses treat marketing as the acquisition engine, not the recurring revenue engine, and I hope that changes.

S5 E110Season & Episode
48 minDuration
14 Stakeholders in a B2B buying group
5 to 8 Years between SaaS buying cycles
1 Day a week saved per rep with AI

"You're certainly not gonna get fired for bringing in Salesforce"

Hear Harry on
The Business of Marketing
Season 5 Episode 110 48 min