This is an AI Transcription. It’s pretty good, but please forgive any errors.
[00:00:00] Jonathan: Welcome to The Difference Engine, the show for tech founders, investors, and innovators.
Okay, Paul, what's coming up today on The Difference Engine?
[00:00:14] Paul: Well, what's coming up is Salesforce as it makes its new case for category leadership again. And our static workforce is a sign of the time. But first, we can't avoid it any longer. Let's go ahead and talk about the dreaded teamwork.
[00:00:29] Jonathan: As you might have guessed, it's the T word.
We're gonna have to talk about him.
[00:00:34] Paul: Alright, so, let's talk about the Trump putt, as they call it. And you'll recall that a putt is a bet that share prices are gonna go down. Uh, otherwise known as the Tech correction, or in our view, a real opportunity for European tech.
[00:00:50] Jonathan: Now, the thing about the word correction is it's, you know, it's in the bag of tech buzzwords, isn't it?
So, it means something, right? Yeah, it does mean something. So, if you're going to talk about a market correction, first thing you've got to talk about is the magnitude. And that is when the decline in the total value of the market falls within the 10 to 20 percent range. Any more than that, and it's a Disaster.
Yes. It could even be a recession. Who knows? Uh, so it's really important. Also, we've got the magnitude. We have to think about the duration. Now, corrections are usually short lived and by short lived, we mean lasting between three or four months on average, which is why we're
[00:01:33] Paul: going to do this right now, because we might be correcting for a while.
[00:01:36] Jonathan: Of course, the frequency of them is they occur more often than market reversals, which my esteemed colleague has just described as big trouble or something like that. And the other thing, which we all know, if you've been around for a few decades in this business, it's about unpredictability and investors.
often don't see corrections coming almost until they're over.
[00:02:00] Paul: Stock markets around the world are struggling to keep up with the noise, by which we're calling what some people have termed the war on attention. Yep. Um, I don't know about you, it's very exhausting to keep up with the stuff coming out from Washington at the moment.
So let's, let's talk about why one would buy a share. Why would I buy a share in NVIDIA? Why would I buy a share in Microsoft? Why would I buy a share in IBM? Well, I'm paying a premium because I'm buying a, a security, a part of a business that is going to give me a better return over time than keeping my net worth in a bank, for sure, uh, in crypto.
Depending what day of the week it is, or even gold, you expect to get a higher return with some degree of safety that most publicly traded companies, obviously with some exceptions, will not go bust overnight.
[00:02:53] Jonathan: Yeah, but of course you have to accept an element. Of risk. Indeed. Which you, dear customer, have to be aware that the value of shares can go up and down, sometimes catastrophically.
[00:03:06] Paul: Whatever your equivalent of a 401k or a pension plan is, or even an employee share option scheme, you're probably seeing a lot of movement. When the premium for owning a share evaporates, Where does that leave people who want to invest in our industry?
[00:03:21] Jonathan: But I mean, more importantly, where does it leave tech investors who take higher risks on unproven techs, um, or immature markets a long, long way before they achieve full value?
[00:03:33] Paul: That's right. So the reason that the tech, uh, shares have boomed so much, we've got the magnificent seven, they were the fangs, God knows what they'll be. Um, with current movements, we'll have to think of a new acronym. But these These shares get a premium even above shares that are in the stock market because they're allegedly high growth.
The investments are based on technology which has not matured and there's a lot of speculation that things are going to get. Way, way better.
[00:04:01] Jonathan: But, you know, we are seeing massive volatility because, you know, what happens, you know, to high net worth investors, private equities, other capital allocators, when value can shift hugely on the basis of a potentially hugely rational or misinformed tweet.
[00:04:17] Paul: Right. So it's one thing to get your information on a quarterly update from a publicly traded company or your investor notes if you're part of a, you're a partner in a VC fund. Right. Um, maybe once a quarter as well, maybe once a year, but it's quite another thing when a tweet, uh, can make everything sort of spiral out of control, uh, and that makes for interesting times, and you'll know that that's the, uh, Chinese curse to live in interesting times, which is obviously, they seem as very bad, but funnily enough, Chinese tech stocks and the Chinese stocks in general are going quite well, so we do live, live in interesting times, and we can see that in everything from food prices, you How much we pay to live somewhere, how much and at what prices our governments can borrow.
Why would you bother about that? You say, well, that's going to make a difference to all of your borrowing rates. So yeah, it does matter a lot. Um, and the correction, uh, has been defined as you rightly says as a decline of 10 percent or more, but less than 20 percent in the price of a security asset or financial market index from its most recent peak.
And when we see the sort of chaos. That we've seen, just seen in the market. It's time to actually take stock a little bit. And you and I have seen a fair bit of chaos.
[00:05:28] Jonathan: Oh yeah, we've lived through a lot of chaos, haven't we? 87's Black Friday, dot com crash. Year 208 global financial crisis. And frankly, many, many more minor corrections in between.
Um, you know, people like to say we've seen this before. I think we have to ask, have we? Could there be something different going on this time? Now, sort of the FT has picked this up a little bit recently. And they did point to a rapid slump in US stocks over the past few weeks. Um, and particularly pointed out.
Dramatic falls for big names such as Nvidia and Tesla, but they think, and I think this is true in most markets when they move below the surface, that's been winners as well as losers. Um, because investors have been switching into stocks viewed as better insulated against economic concerns and some of the cratering
[00:06:25] Paul: And like, and so I hear.
That's, uh, old school tech stocks from a tech perspective like IBM and Cisco, who would have thunk it? What does this mean from a category perspective? Well, you know, we love tech and tech is where the categories of tomorrow are normally built, but it's a sector that is more sensitive to ups and downs than most.
It's also a leading indicator, meaning, uh, if a recession is going to happen, uh, and we start to suffer from corrections. It's often the first one in, and also the first one out, and it's the first signal of recovery. But we are very far away from that at the moment, because we are not in normal times, and we don't have normal, market driven economic cycles.
[00:07:04] Jonathan: These are the days of the dominance of the Magnificent Seven, and their massive and disproportionate valuations. So perhaps Just perhaps this correction will hurt those guys more than the others. You know, could it be that Adam Smith's invisible hand is visibly delivering the conditions for the next episode?
[00:07:21] Paul: Certainly. If you look at a chart and there's one in the FT, we can stick in the show notes of the, um, mag seven, big tech, uh, the sell off in recent weeks has been. Tremendous. So betting against big tech
[00:07:32] Jonathan: has actually paid off in recent weeks.
[00:07:35] Paul: Is that the Trump putt? I don't know. It could be. And just when all of those guys cuddled up to, uh, to Mr.
Trump, how ironic. Now category designers and would be category designers. Knock, knock. Opportunity knocks. It certainly does. So, you
[00:07:48] Jonathan: know, wanna build a rival to US defence systems like Patriot, F 35 or Starlink?
[00:07:53] Paul: Now's the time! Yeah, I mean, it might seem like an impossible dream, but, uh, if the option is no Patriot air defence, no F 35s in the sky, and Elon pulling his Starlinks, as he's, um, threatened to do to Italy and several other places, Now is the time.
If they don't pull Starlink first. You need to build a better data sovereignty into your large language model. Speak up. Why would you, um, decide to give all your data to somebody that wasn't playing nice?
[00:08:18] Jonathan: Yeah, and if you're able to use friends and family to avoid being told what to do by a spook tech investor, get
[00:08:24] Paul: your bootstraps on.
Get your bootstraps on. To be fair, it's not just us calling this. Some of the more enlightened in the media, and we do love, uh, John Thornhill, the sifted founder. He's already calling this out. His piece, uh, his opinion piece again in the hallowed financial times is it's time for Europe's deep tech companies to go big.
But John has a little bit of skin in the game, doesn't he? You
[00:08:46] Jonathan: know, his baby sifted is a frankly, a business predicated. on the premise that European tech founders are still being funded.
[00:08:55] Paul: Yeah, and he's also talking specifically about deep tech, right? Not SaaS, which we know may be about to be blown up and fragmented.
Um, not lookalikey apps, not rip offs of things that are already out there and certainly not lifestyle tech. He's talking about fundamental deep tech, something that You know, Europe could probably do a better job of it. He served up a really
[00:09:16] Jonathan: nice quote, didn't he, in the FT? You know, with the political winds blowing on their backs rather than in their faces, it is time for Europe's founders to go big.
They will now have to acquire new skills in managing rapid expansion.
[00:09:30] Paul: Yeah, so as the cliche goes, uh, and it is very American, the nexus between markets, global politics, And tech is right with us. And if there's one piece of advice we can share, it's simply focus on the tech and build big categories, do that.
And God willing, the rest should take care of itself. Maybe, uh, turn off the news alerts for now. So, you know, we know markets
[00:09:54] Jonathan: oscillate. Maybe never as frequently as today in the tech sector. Um, we've seen conditions we've never before seen in a comparatively young tech economy. When an ill judged or malicious social media post can send us stocks spiraling.
But remember, tech is not going anywhere. Questioning the value of the overblown Magnificent 7 could be just the thing to stimulate new categories. And it's not just that. You know, Europe's Europe, you know, has every incentive to compete. Tariffs, not to mention ridicule and exclusion from U. S. markets. LFG, European category creators.
We think that Salesforce is making a new case for category leadership. Now, as we predicted in a recent episode of the Difference Engine, number 43, in fact, the category era of SaaS, as we have known it, could be coming to an end. Well, why? Well, we think the software industry will start rapidly pivoting to, uh, I'm going to call it triple A S there are other pronunciations are available.
As my glamorous, uh, colleague has just indicated that's agents as a service. If you've not been paying attention now. It looks like Salesforce has truly broken cover with its execs at TDX25 developer conference in San Francisco in recent days, extolling AgentForce as a whole system AI play. That means an agentic AI technology approach, not hung up on large language models, LLMs, agent force, they say is the Holy Trinity should appeal to you being a nice Catholic boy, Paul of data apps and agents.
Now, apparently there's also a foundation model, and there's a series of smaller models that go into that Atlas reasoning engine. And they're automated workflows that people can draw on so they get it. Everybody's been talking about AI models over the past few years, but in reality, we need to be talking about systems that can actually make a difference in the workplace.
[00:12:04] Paul: Okay. Well, I mean, you have to give it to Mr. Benioff. He really does know how to mold a category. And by the way, agent force, despite the glamorous, uh, Matthew McConaughey adverts that we're seeing all over the place, this is not their first rodeo by far. Um, they've been in agentless or, or AI, let's say, for a very long time.
Uh, there's been a lot of trial and error. You'll remember the magnificently titled product Einstein, um, which then morphed into Einstein Copilot and a million different other varieties as they seem to duck and weave and try and find out which is the product that's going to give them the category fit for this.
So for the category curious, um, notice the repositioning. So Einstein GPT, as it was styled. Back in the days, probably four or five years old, and there have been various iterations of Einstein as they use open AI as a co pilot and then Microsoft as a co pilot, et cetera, et cetera. And the recent, the first use case that they found for Einstein was, of course.
Sales automation. Well, Salesforce is sales automation, and we happen to know that that is a use case that has seen probably the most amount of A. I. Goodness thrown at it now with the world, the world moving to a gentic A. I. Those simple use cases are becoming a bit obvious. So now where are we, Jonathan?
[00:13:27] Jonathan: Well, you know, I think this is serious. So they're really going for it and setting the pace. You know, the entire Salesforce technology stack, including slack, MuleSoft. And Tableau comes into play as a system with an agent force, as does its Data Cloud platform. Now, essentially, they're positioning, I think, Data Cloud as an operating system for AI.
It's about potentially harmonizing hundreds of thousands of systems containing the data, the metadata. And the semantics now, they are claiming that the data cloud makes LLMs actually useful by eliminating the things that hold them back in real life. And we know what they are, hallucinations, bias, toxicity, and a whole lot more.
So they're adding to the LLM on the basis that the data pairs, the AI, then the AI, AI powers the customer experience. It isn't going to stop at the data clouds. They have their sales cloud. Service cloud and marketing cloud apps, as well as the aforementioned Slack, MuleSoft
[00:14:30] Paul: and Tableau. You have to wonder at which point the tech stops and the marketing starts with this, with these guys.
[00:14:35] Jonathan: Well, they're already making some interesting marketing claims. As we said, they're really going for this. Um, so they've got AgentForce And they claim that they could turn on 10, 500 customers over three days with agents. And we know why that is. That's because they're using the same platform as they have for the last 25 years.
Now, that could make a lot of sense for any enterprise customer wanting to evolve their applications or indeed the whole organization. If they have the workflows built in their Salesforce deployment, those could all be made available for virtual agents. So I think this is a big growing up moment. Too many organizations are doing what Salesforce has termed, uh, DIY AI.
[00:15:17] Paul: And you better get ready for a lot of categories that've got something. AI voted on on the front.
[00:15:22] Jonathan: Yeah, but we're thinking this DIY AI is going to be a slogan that we'll be hearing a lot more of in coming months, just like the iconic, iconic, no software, right? So their definition of DIY AI means customers are taking whatever enterprise apps they have, you know, whether it's Salesforce or SAP or Workday.
buying chat GPT or another LLM and trying to plug it in. You know, the difference is Salesforce is not shipping just another model or app or a co pilot. They are shipping an AI system on an already deeply unified. And clearly, very widely used platform.
[00:16:02] Paul: Well, that would be helpful because what we see out there is a lot of people trialing, as you say, DIY, AI, DIY, LLMs, specifically variants of the open source DeepSeq, which brings with it a hell of a lot of problems around data sovereignty.
Data privacy, IP theft, et cetera, et cetera.
[00:16:19] Jonathan: Yeah. So that thing we always have to remember in, in a category, it's all about what the customer wants and the customer never wants complexity. And the customer really doesn't care how it works. They just want to know what does it do? So. Salesforce are running very hard at this, uh, and rightly so in the absence of any serious competition.
They are actually claiming to have already sold this to over 5, 000 customers in the last six months. Now, by my reckoning, that makes only another just under 200, 000 customers to go. And with that, of course, an exceptional opportunity to rapidly build a AAAS. Category lead,
[00:16:56] Paul: if they don't all morph off and do their own thing, a word of caution here about our friends at Salesforce who are legendary at both categorization and marketing.
It's the details that matter when it comes to this new world of, um, it
[00:17:07] Jonathan: is
[00:17:08] Paul: a S or whatever we're calling it. Um, and the interconnections of the systems underneath all the glamorous. Front end that we see is really important. These days, a lot of the whizzy e commerce systems that we're using still have to use very archaic, I'm talking from the 70s, EDI systems.
Like banking. Yeah, banking, e commerce, retail etc. Now, Salesforce does have its own. You mentioned their API management tool, MuleSoft, but that's only one of many, right? Not every enterprise, even if they're quote unquote, Salesforce end to end, has, is going to go for one API management. They just can't, with these supply chains today, you have to link to whatever your suppliers want to link you to.
So, um, AgentForce is going to have to live with going out to source data from other parts of the world. World, which are not part of the salesforce ecosystem on. It's going to be very interesting how that all works, but what it does do for all of those salesforce partners, and there are many thousands of them in the world, is it?
I smell a good opportunity for a lot of a I connect them, make it all work type services.
[00:18:21] Jonathan: Yeah, well, you know, that's that's undoubtedly true, because what? Salesforce is doing with this move is potentially creating a brand new ecosystem. And, you know, let's face it. Salesforce has been an absolute leader in creating very, very efficient, massively value add ecosystems.
So you have to ask yourself, why shouldn't customers go for this? Rapid evolutionary approach. I mean, as we talked about in episode three of the difference engine, this sort of enterprise agentic approach is one where organizations can use the resources they have and add new whole layers of value to them.
You know, I mean, we're talking about here, everything from. Transactions to engagement and other data workflows, existing automation, and of course, even AI policies. So the category genius behind all of this is that it's a wholesale alternative to the LLM wars and generative AI copilot usage. Um, and of course, who's pushing those two particular areas?
Well, it's Microsoft and Google again.
[00:19:27] Paul: Right, and Microsoft and Google would love to compete with some of the other properties that Salesforce has got, you know, on the Tableau front, you've got Google Looker. And of course, Salesforce doesn't own the layer below the applications. It doesn't have as far as far as I'm aware, massive data center capabilities.
So Yeah, they've got a big old fight on their hands.
[00:19:47] Jonathan: Now we've said previously that Google and Microsoft could own horizontal desktop AI space. You're doing that through the use of co pilots and LLM based search. It likely owned a desktop productivity because they have the distribution. But equally, there is a huge category opportunity.
In the vertical markets for new AI category players to flourish, but even with billions being invested in co pilots, they're pretty tactical, we think, and we wonder just how much productivity increase they will actually deliver. Yet here is Salesforce using its existing ecosystem and building on its existing assets to capture a load of vertical markets and some of the grown up horizontal enterprise plays too.
[00:20:30] Paul: Yeah, because all of this capability, let's remember an agent force or whatever. Is built on co pilots. So they're just owning the front end, which is exactly where they should be and exactly where they've succeeded and excelled in the past.
[00:20:42] Jonathan: Yeah, and I'm sure Salesforce is very happy with that because they can put the value into the back end.
And I think, I think this is genuinely more Mark Benioff genius. Um, but this time for the age of AI, um, we think. ushering in a new age of category leadership. I mean, it certainly looks that way.
[00:21:01] Paul: Well, thankfully, um, we have got some more goodness coming the way of our listeners, right? Yeah,
[00:21:06] Jonathan: if you're wondering how Mr.
Benioff has been able to succeed where other leaders in the Oracle Diaspora failed. Make a note to tune into our upcoming episode of the Difference Engine. Then our special guest in association with BoardWave will be none other than Dr. Steve Garnett, the man who built Salesforce in EMEA, as well as Siebel and Oracle, and we'll be covering that too.
[00:21:31] Paul: Right. So mark your calendar, bring your notebook and maybe put those notes. into a generative AI.
[00:21:45] Jonathan: You know what? This really
[00:21:47] Paul: grinds my gears.
Okay listeners, it's time for Grinds My Gears. Let's vent some frustration, Paul. Indeed, I'll tell you what's been grinding my gears, and it's maybe a sign of the times, is Static workforces.
[00:22:01] Jonathan: Static workforces. Does that mean that they're sort of covered in an electricity that crackles when you go into the office?
I think
[00:22:07] Paul: it just means hiring and firing, actually. But there's a lot of it going around. Tech unemployment, I mean. And if you don't know any former colleagues or even current colleagues who are looking, um You're living under a rock. Yeah, where are you? Weirdly though, the numbers don't really play that out, but we'll get to that in just a second.
Um So yeah, the tech layoffs are generally getting less attention than they did in the full on tech lash of 21 through, let's say 23. But we are in an era of what some are calling static workforces. That's a
[00:22:35] Jonathan: strange euphemism, Paul, isn't it? Because we're actually talking about shrinking headcount. We
[00:22:39] Paul: are.
Nothing static about it. We are. And the size of it. So our friends at Positive are Currently hiring for a number of tech marketing roles and have received an avalanche of interest money who've left for a recently, very recently, um, last week, in fact, in some cases for much larger firms, um, who are looking to, um, put their wage bill through the exemplary treatment.
Other weight loss drugs are available. And it's not just the, um, those who help tech right in the heart of, you know, there's the sweethearts of the tech and the UK, uh, FinTech space, uh, for instance, or friends that go. Cardless. These scale ups are looking for profitability now because they're smelling an IPO or an exit.
It's not just
[00:23:19] Jonathan: the start ups, you know, mature firms like KPMG are doing wholesale layoffs.
[00:23:23] Paul: I mean, GoCardless had a 20 percent riff, which was, which was hard for a smaller company, I'm sure. Uh, although the nice chaps there, and shout out to Hiroki and the crew, did vest everyone, which is nice of them, who exited, and added two months pay to their statutory entitlements.
So, hopefully, the diaspora will be kind when they come back with their Revenge startups this year even the big guys even the Goliaths of tech who have hired Armies of engineers to build product and sales teams to sell them are slimming down I'm not saying say I'm not saying Mark Benioff's slimming down that might arguably be a good thing.
That's up to mark That's up to mark, but he um definitely declared in a recent survey Core that he would not be hiring more engineering talent this year. Where have you ever heard that?
[00:24:04] Jonathan: Well, it's also particularly interesting given their great big
[00:24:07] Paul: AI push, which we've talked about earlier in episode.
Yeah. And maybe they've had enough people to build that product and how they've sort of, uh, hoist themselves by their own petard, so to speak, uh, Microsoft. Another, um, one of the, uh, in fact, uh, definitely part of the Magnificent Seven has, um, let 10 percent of its workforce go and they've been a little bit, uh, snidey, uh, shall we say about it?
Um, they've let these people go to apply their wares elsewhere, but they've cited Performance as a reason, which you could argue is a little bit of a nasty dig for anybody who's looking for a new job.
[00:24:46] Jonathan: Honestly, I think that's the right way to go, because if you look at tech cycles over and over again, as you go through the boom period, firms suck up an awful lot of people.
And I would say that they generally do not do sufficient due diligence on those people. And then when times get tough, they find that a lot of the people of hard are really lacking and the worst thing you can do is to keep underperformers in your organization because frankly that is corrosive to the people that do perform
[00:25:15] Paul: right and I know this is a hot topic for yours because you're very good at uh at the recruitment side of things but here's another perspective from um the sort of bitch fest uh that is um Glassdoor and here's somebody saying that Microsoft has laid off 2, 000 underperforming employees Many of these people 10, 20, 30 years in the company, which means they get a decent severance package, which would help them keep their families employed for a few months and while they find a new job, but given Microsoft deemed them as underperformers, have they been black labelled?
For a long time, um, and where they fired without warning, uh, and no severance and all pay and benefits dropped, uh, following that, which is what the accusation was. Well,
[00:26:03] Jonathan: I'm sure the truth will eventually emerge on that, but big tech firms. Even quasi big tech firms like Amazon are letting people go and taking a long, hard look at them.
[00:26:14] Paul: They are. And I wonder if they're also keeping it quiet. Because a quick look at layoffs. fi, which was beloved of the media when they were having a good old tech lash, says, according to the numbers, That the layoffs are not that bad. So just to give you those, there were 264, uh, thousand people laid off in 2023 in tech, uh, about just, uh, slightly more than half that, 152,000 in, uh, 24, and so far a third of the way through this year, only according to layoffs.
22, 000 or 23, 000 people laid off, which would put us on track for fewer layoffs this year than in the previous years. But I just wonder if that's the full story. And if there's a little bit of, um, hiding the numbers, not being very explicit about why people are going.
[00:27:03] Jonathan: Uh, just about levels of normal staff turnover in the tech industry.
Uh, just this time, uh, management are actually getting in charge of the turnover rather than the employees.
[00:27:13] Paul: Yeah. So I think it's, so what, what are we seeing going on? Is this, as you say, just the usual tech boom and bust cycle? Uh, is it a sign that the old guard employers at Microsoft Salesforce are mature and then need less talent for whichever reason?
Is it greedy shareholders of those companies saying, Ooh, um, if we slim down our workforce, we'll make. tastier profits. Are they, are these companies battering down the hatches for whatever happens from a recession point of view?
[00:27:39] Jonathan: It's battering a more extreme form of battening then.
[00:27:43] Paul: So we're really,
[00:27:46] Jonathan: really putting the hatches down
[00:27:47] Paul: with maximum prejudice.
You're also going to whack on the head while you're doing it. Yes. Thank you for that. Uh, or is this the payoff from AI coming? Um, yeah, don't know. Cause we hear from other quarters, specifically a lot of commentators in the States that static workforce is a thing. This means. According to the definition, smaller or the same headcount achieving what a pre layoff team could do.
Back to your performance thing, right? So, so
[00:28:10] Jonathan: is this actually tech doing what it's been preaching to all those companies it's been selling to over the time? Right? You know, tech will give you the Nirvana of efficiency. And of course, AI is going to take it to another level.
[00:28:23] Paul: Well, right, because as you say, coders are finding today, uh, huge productivity gains from co pilots.
Um, but the static headcount philosophy might find out that those gains get them a new career, so to speak, as they're exited from the company.
[00:28:38] Jonathan: Yep. Everybody has to be prepared to have a whole series of careers during their working lifetime. And actually, those careers actually are looking increasingly like a series of projects.
Gone are the days when you have a career which is climbing the greasy pole in probably one or two major corporations. That is over. Just
[00:28:57] Paul: a bit of advice for those following this static head count. Um, meme to its logical conclusion, maybe a word of warning for me and maybe for my colleague here, a word of encouragement.
Um, think about lay, what happens if you lay off the wrong talent, could they come back and bite your backside, uh, as competition? Highly likely in tech. Um, and even those who do stay, uh, the guys that you didn't riff, uh, will they resent their former colleagues and workmates being fired and edged out? And will they find that they've got even more work, even with the help of AI assistance as much as it can provide?
Will they find that they are doing more work and naturally hold a grudge?
[00:29:41] Jonathan: Yeah, so I would say higher slowly. And really, really understand what you are hiring for and what the capabilities are and what the attitudes are in the individuals you're hiring. And you really do need to give a damn on the way in, because you're going to have to make some painful decisions on the way
[00:30:00] Paul: out.
[00:30:04] Jonathan: Thank you for listening. If you want to learn more about category design, head to Becategorical. com. If you need help designing and dominating your category, then get in touch. Contact details are in the show
notes.