Latest / State of the AI Union / Steve Baker: Where AI Deals Are Won or Lost
Transcript
- 0:06Hello everybody and welcome to the new year and this week's
- 0:09episode of the State of the AI Union.
- 0:11Today I'm joined by Steve Baker, who is Co founder and CEO of
- 0:16Vendor Sage. Steve, thank you for joining.
- 0:19He's coming all the way from New Zealand.
- 0:22I'm so very excited to have him. Steve, would you like to
- 0:25introduce yourself and vendor Sage?
- 0:27Yeah, for sure. And thanks for having me, Laura.
- 0:29I really do appreciate it. One of the Co founders of
- 0:32Intersage and the way we like to view ourselves as the execution
- 0:35layer between business intent and software reality.
- 0:39So we started off as a SAS management and procurement
- 0:42platform and we've shifted more into tech strategy execution as
- 0:45a copilot as well. So to clarify, that means that
- 0:52the this copilot helps us understand like usage of all of
- 0:56the tech that an organization has procured and is currently
- 1:00using. And you can ask questions about
- 1:03usage, adoption, that kind of thing.
- 1:05Yeah. So the the way I like to, the
- 1:08analogy I like to use is a business will have goals it's
- 1:10trying to achieve. It'll hire people to achieve
- 1:13those goals, give them job descriptions, KPIs, and then
- 1:16it'll manage them. We do that, but for the software
- 1:19you use to run your business. So we're continuously looking at
- 1:22the capability of the tools, the health of the workflows.
- 1:25And then off the back of that, what we're trying to do is
- 1:27figure out how to bridge the gap between the current set of
- 1:29capabilities to wherever it is it needs to be to achieve
- 1:33business goals. So think of it almost as a set
- 1:37of eyes and ears on everything that's happening inside your
- 1:39tech stack and where you can take it from to where it needs
- 1:42to be. And that's actually like one of
- 1:46the things that's like very important right now as we see
- 1:51tech stacks are consolidating, there is an increased pressure
- 1:54from, you know, CF OS and finance teams to reduce cost,
- 1:58consolidate, etcetera. So measuring and being able to
- 2:03kind of see where our money is going and whether we're getting
- 2:06the returns is, is really interesting, I think for me.
- 2:10Passively, yeah. You also mentioned that this is
- 2:13the execution layer between goals and software.
- 2:16And I wanted to ask, you know, now that companies are adding
- 2:22more AI tools, why does that strategy to execution get become
- 2:28wider instead of smaller? It's a good question.
- 2:31I think to like pre face that it's important to understand the
- 2:35problem, which is that businesses still fundamentally
- 2:38don't really understand how to do a good job of buying,
- 2:41managing and adopting software. We surface business challenges
- 2:46and then we we really good at picking out tools.
- 2:50We don't really know what happens off the back of that.
- 2:52And in the nature of like how I look at software companies, if
- 2:57you think of like your product road map, there is a cohesive
- 3:00direction that everything is going.
- 3:02You, you wouldn't have a product owner unaware of where certain
- 3:06parts of the feature of road map are going, but without tech
- 3:10stacks, because the tech stack is basically you can view it
- 3:12like a product. You to to not know what
- 3:17direction certain parts of the text that are pulling in is.
- 3:19It's, it's a pretty dangerous space to be in, especially when
- 3:22you're looking at how fast we're rolling out use cases across the
- 3:24Thai olds at the moment. And so I think like the way I'd
- 3:28like to think about it is the AI tools don't really shrink,
- 3:30shrink the strategy to execution gap.
- 3:32They just multiply decisions. So, you know, class success was
- 3:37really easy to govern because it had clean signals, licenses,
- 3:40adoption. The renewals are really
- 3:42straightforward. But with AI you get tools
- 3:44popping up everywhere writing research support, sales code,
- 3:49and each one of those changes of workflow in a slightly different
- 3:51way. And the governance on the top of
- 3:53that is, is very much not in place at the moment.
- 3:56There's no framework for execution and governance in any
- 4:00real sort of capacity. It just normally ends up being
- 4:02led by one or two champions in the org, but it's sort of just
- 4:07have that at the moment, which I think is leading to a lot of
- 4:10challenges in adoption. I think there was a step that I
- 4:12saw the other day that was 90% of AI initiatives fail.
- 4:17SAS set historically at about 70%.
- 4:21And so 70% of software initiatives or implementations
- 4:24didn't end up doing what they were supposed to do.
- 4:26And we still haven't learned from that.
- 4:28And it's compounded into AI, which is actually just a
- 4:30distribution mechanism in most cases.
- 4:34Yeah. And so we're still seeing the
- 4:35same problems. It's just it's a lot easier to
- 4:38adopt and get it in the hands of people, I think is what's
- 4:41changed. But we're focused very heavily
- 4:46on the how much Christian at the moment.
- 4:49We're trying to figure out how much things cost for how we're
- 4:52going to manage the spend. And that's not really the right
- 4:54question to be asking from my perspective.
- 4:57I think the right question to be asking was, can we get it to do
- 4:59what it's supposed to do? But yeah, most teams don't
- 5:03really know how to get to that part, so they default to what
- 5:05they know how to manage, which is spent.
- 5:08Your spend only goes out of control once, once you've really
- 5:13ended up at a place where there hasn't been any sort of
- 5:15thoughtful focus on how you build the stack in a cohesive
- 5:20way, I guess. Yeah.
- 5:22You know, you said something a couple of things really
- 5:24interesting to me. The first one that you said was
- 5:28AI actually just gives you more decisions to make.
- 5:32And I think that that's kind of fundamentally the, the, the
- 5:40difference between doing AI right and doing AI without
- 5:44direction, right? It's you want to, you want to
- 5:48use AI in a way that is actually changing the way that you're
- 5:51doing things so that it's making it easier.
- 5:54But if you're deploying it in a way that's not thoughtful, then
- 5:58you actually end up with multiple ways that you have to
- 6:01look at it more decisions to make make.
- 6:05And This is why those AI projects fail, because we
- 6:08actually haven't thought about or been thoughtful about like,
- 6:11well, what do we want it to accomplish and how are we going
- 6:14to do things differently to be able to get that done?
- 6:19Yeah, same problem with SAS, right?
- 6:22The different, the difference is with AI, sales reps can now say,
- 6:24what do you want it to do? They're still just looking for a
- 6:27way to get you to say yes to something, right?
- 6:30And so we we get to see how some of the fastest growing companies
- 6:33in the world make the choices they do about the software they
- 6:36use to run their business. And the alarming thing for us a
- 6:39lot of the time is when they're doing these cross company
- 6:42evaluations for like open AI or oceans agent or those sorts of
- 6:47things. What we're seeing is one or two
- 6:50roughly designed use cases like that.
- 6:53That's the full extent of where a lot of like some of the
- 6:55fastest growing companies in the world is deciding, yeah, we'll
- 6:58go and figure out the rest from there.
- 6:59So there's no, there's no understanding of what good looks
- 7:03like, what success means once it's where it's supposed to be.
- 7:06There's no plan to roll out and implement, no ownership or
- 7:09anything like that. And that's the problem we had
- 7:11with SAS. It's just it's moving twice as
- 7:14fast or three times faster to get to roll out with, with AI.
- 7:18And I think that's sort of what's compounding those issues.
- 7:22Yeah. And, and also you mentioned
- 7:25that, you know, spend only becomes an issue when actually
- 7:29people start adopting it. And with AI, you're, you are
- 7:32supposed to use more AI that, that is kind of like success,
- 7:37but then it can go out of control.
- 7:40So actually this is my second question here.
- 7:42It's, it's around how, how do finance and IT teams I guess
- 7:49more on the finance side like how did they now evaluate
- 7:52software with this AI lens and it being everywhere?
- 7:59I think this is the, this is the broken part.
- 8:01So finance and IT are typically they try to position themselves
- 8:06in forward thinking orbs as enablers, but they're not,
- 8:08they're the gatekeepers to budget into implementation,
- 8:12right? What's really changed with AI is
- 8:15you, you actually have to start to break the silos down to
- 8:17really understand what's happening.
- 8:20Before you could say we're just going to implement the CRM.
- 8:22It's going to do these five different things.
- 8:24Now what's happening because of the inability to go like, I can
- 8:30only make this agent do this one thing.
- 8:32You end up deploying a ChatGPT as an example, and you blink and
- 8:36within within a week or two, you've got 2 or 300 different
- 8:39new workflows deployed. So what we're what we're seeing
- 8:45is a shift is a much more collaborative evaluation cycle
- 8:51taking place where it's not just sign off from finance and IT
- 8:55there's a real careful look at what is truly being enabled and
- 9:00what good is going to look like. So yeah, I think the, the the
- 9:05ones that are doing a good job are just taking that extra bit
- 9:07of time, not really leading into the hype as much that's not
- 9:12being what as worried about the FOMO piece and going through
- 9:15much more careful evaluations. I think the last thing I call
- 9:19it, as well as where we're seeing successful deployments of
- 9:24AI, is when there are proofs of concept completed first.
- 9:28So predefine what you're trying to achieve, go and test it, roll
- 9:32it out and see if you can actually do it, measure it if
- 9:36you want. Go through another Pfc again at
- 9:37a scale that phase, and then off the back of that, you've
- 9:40probably got a good basis to make a smarter choice about what
- 9:42you're doing. Who do you think should lead
- 9:45that EA initiative for company wide type of thing?
- 9:51Is it? Is it?
- 9:52I actually think it should be a chief of staff, right?
- 9:55So what you what, what we're trying to really do with AI, the
- 9:59point of it is efficiency and we're trying to operationalize
- 10:02things, make them move faster. Workflow improvement really is
- 10:06what it's, it's changed everything in that respect.
- 10:10So working backwards from strategy, bridging the gap
- 10:13between what are we actually trying to achieve as a business
- 10:16this year, what are the capabilities we need to achieve
- 10:20that? And then what is the current
- 10:22workflow house and what do we, what happens if we improved
- 10:26workflow health from X to X? Does that move us towards where
- 10:30we're going faster? What you don't want to do is end
- 10:32up in a silo making a choice. You need to have visibility over
- 10:36how information moves through through the entire business.
- 10:39So. Someone that's like really cross
- 10:41functional has a good idea of, you know, connecting the dots
- 10:44across the business, that kind of person.
- 10:47OK, Yeah, where where we used to measure ROI for software was
- 10:53really in the in licensing seats and adoption.
- 10:57And then we moved to consumption, you know, more
- 11:00recently in the last few years. But now AI and having these bull
- 11:07use cases and a lot of them are also, I guess, I guess in my
- 11:13opinion, you know, AI is supposed to just help us become
- 11:15more efficient. When when that value is not as
- 11:21direct, how should companies think about measuring it?
- 11:26And how does like vendor sits like help with that?
- 11:31I think it's it's really difficult to think of AI like a
- 11:34product. It's probably measured more like
- 11:38an intervention. So you think inputs, outputs,
- 11:41outcomes, second order effects. But SAS adoption was often a
- 11:46proxy for value. With AI adoption can be
- 11:49misleading. So if you think about, you know,
- 11:53there being 100 different people active this week on your ChatGPT
- 11:58instance, that doesn't necessarily mean that they're
- 12:00doing what they're supposed to be doing.
- 12:03Whereas with your CRM, if you saw 150 calls logged, 180
- 12:07support tickets completed, so on so forth, those were those are
- 12:10good outcomes. So yeah, like someone can
- 12:14generate a lot of activity without actually creating better
- 12:16outcomes. And AI is really good at that.
- 12:18You can get distracted really easily because it can do
- 12:20everything. So when when we talk about what
- 12:24good looks like, what our, our copilot actually really prompts
- 12:28and forces people to move towards is figuring out what is
- 12:33what is, what is the one thing, what's the thing that's going to
- 12:35move the needle? How are we going to improve this
- 12:37workflow's health? And how do we shape and mold
- 12:40your tech stack? The parts of AI and the parts of
- 12:43SAS that are relevant to those workflows to getting healthier
- 12:48and ultimately contributing towards capabilities that enable
- 12:51businesses to achieve their goals.
- 12:54So what I'm hearing actually is that we should measure the.
- 12:58We should measure the value of AI by the business outcomes that
- 13:02we are already measuring and whether we're doing more or less
- 13:06same than what we were doing before the adoption of the.
- 13:10Yeah, we, we, we encourage companies to pick one workflow
- 13:15at a time and measure it sort of end to end.
- 13:17So your inputs, your usage signals just enough like just to
- 13:21normalize it. Then from there, your outputs,
- 13:24your drafts, your suggestions, your summaries, outcomes off the
- 13:28back of that cycle, time resolution rate, conversion
- 13:30rates, defect rates, how you would manage a person.
- 13:33But just started to think more of AI as a person, I guess.
- 13:38Yeah, it's, it's a very different way, a very different
- 13:42way to approach it than what you'd be doing, I guess with
- 13:45SAS. I like what you said, think of
- 13:48AI as a person because I never, we always think about, you know,
- 13:51AI as a teammate. We try to use, you know, AI as a
- 13:53teammate sort of language. And I think you're echoing
- 13:57exactly that. Yeah.
- 13:58And like that example I gave at the start, right.
- 14:01I actually do view software as as a part of the team.
- 14:06I always use the analogy of, you know, what every single person
- 14:10in your business is supposed to be doing.
- 14:12I have a job description. They have KB IS you manage them,
- 14:15you work them towards, you know, improving on what they should be
- 14:19doing for the business. You should be doing that with
- 14:22all software, but especially with AI.
- 14:24It requires constant tweaking, adjustments and making sure it's
- 14:27moving towards where it's supposed to be going.
- 14:30Right, right. OK.
- 14:31Living. Breathing.
- 14:33Last question and then we'll get to like some fun, fun questions
- 14:36here. What what advice I think would
- 14:40you give to sellers that are trying to sell AI right now and
- 14:46how would you also position them for maybe having that continued
- 14:51success story since 90 of the implementations are failing
- 14:56right now? This is the same thing that
- 14:58happened with SAS. I keep going back to that
- 15:02example, but the best reps in the world, and I probably am a
- 15:07part of 1000 plus evaluations a year, the best reps in the world
- 15:11are the ones that really do get obsessed over the success.
- 15:16You'll see those reps go on to become enterprise reps for the
- 15:18likes of your sales forces and open a eyes and and so on so
- 15:23forth. They get really, really obsessed
- 15:24about their customers outcomes. And to do that, it means you
- 15:27have to understand your customer, you have to understand
- 15:29the workflows. And it's no different with AI at
- 15:33all. I think a lot of reps are
- 15:36making, making a lot of money at the moment because they're
- 15:40banking on the fact that, you know, whoever they're selling to
- 15:43has been told, I need you to invest in AI.
- 15:46And that's it. Like that's, that's the full
- 15:48brief. But that's contributing to, you
- 15:52know, fears around this bubble popping because we're seeing
- 15:54these churn numbers that we're seeing.
- 15:56We're seeing these fail rates that we're seeing.
- 15:58But there's this very small circle that are getting it
- 16:00right. And that actually does start
- 16:02with a very, very thoughtful vendor relationship.
- 16:07I believe there's vendors that are taking very careful care to
- 16:10make sure that things are rolled out correctly rather than just
- 16:12growing at all costs. And those use cases that we're
- 16:16seeing deployed across our customers are incredible.
- 16:18Like it's completely revolutionizing new businesses.
- 16:22But yeah, it typically does does start with the Rep just taking
- 16:26that extra bit of time to get to know and understand the
- 16:29customers businesses. Can you give us an example
- 16:32actually of a customer that's doing this really well?
- 16:37In terms of the the reps doing that really well.
- 16:40The OR the use case of of what what they're doing really well
- 16:43and that's sticking for them. Yeah, yeah, for sure.
- 16:46So I, I think the, the ones that I've seen really get this right
- 16:50have probably been clean. Yeah.
- 16:54So we, we, we saw that the team at clean and a couple of
- 16:56different evaluations take the time to slow down quite a bit.
- 17:01Really, really understand like where what would move the needle
- 17:04correctly and glean has the ability to do endless amounts
- 17:09right. But what they, what they really
- 17:12focused on trying to understand was if you look at the
- 17:15capabilities required to move the business towards where it's
- 17:18trying to get to at the moment, what are the, what are the three
- 17:21or four workflows that if we, you know, 10X, what are those?
- 17:26How do we, how do we show quick wins there?
- 17:28How do we actually prove out that that is what we are capable
- 17:32of delivering on really easily in Apoc?
- 17:35Get that done, get that done right up front.
- 17:38It removes any objections really quickly and it builds a lot of
- 17:42trust right up front so that you typically see, you know, a lot
- 17:47of hesitation and caution towards going big right at the
- 17:50beginning, especially with SAS. But yeah, there's bigger, bigger
- 17:54full cycle deals from an AI rollout perspective.
- 17:58They typically start small with a couple of really key ones like
- 18:01that. Awesome.
- 18:04All right, Steve, thank you so much.
- 18:05Let's pivot to learning a little bit more about you.
- 18:08And you're from New Zealand, which city?
- 18:10New Zealand, by the way. Auckland, so right at the top.
- 18:14OK, I've been there one time in my life.
- 18:18But I think one, one question, you know, we, we don't see as
- 18:21many founders from New Zealand. And I wanted to ask, having
- 18:26grown up in New Zealand, how do you think this has shaped the
- 18:30way that you think about leadership or business,
- 18:33especially compared to, you know, Silicon Valley culture?
- 18:36I know some of your, you know, backers are out here in the
- 18:39Valley. Yeah, yeah, for sure.
- 18:42We're in in New Zealand. I think we we tend to do things
- 18:47a little bit slower pace than what the rest of the world might
- 18:49typically do or might be perceived to be that way.
- 18:53I definitely think New Zealand is a lot more cautious and less
- 18:58risk adverse. Whenever I have had the
- 19:01privilege of working with companies based over the states,
- 19:04you see a mentality that enables people to put their hand up to
- 19:10do anything. All right, there's very much
- 19:12epic me give me the chance to prove myself.
- 19:14And that doesn't really happen as much here in New Zealand.
- 19:17So for me, that was an opportunity.
- 19:20There's a lot of opportunities to do that, to be able to put
- 19:23your hand up so much more so than you probably would get
- 19:25anywhere else in the world because it's just not as
- 19:27competitive here. Yeah.
- 19:29And, and I think that in terms of the leadership style that
- 19:32I've been able to, to build, this has really allowed me to
- 19:35focus on, on seeing the opportunities for, for each and
- 19:40every part of my team taking a little, a little bit more time
- 19:44just to to understand where the opportunities sort of sit.
- 19:48And yeah, I think, I think what's really cool about growing
- 19:52up here is you, you learn that there, there are a lot of
- 19:58different ways that things can be done.
- 20:00You know, we, we have this saying in New Zealand called #8
- 20:03wire mentality, which is basically we invented #8 wire
- 20:08down here, which is like a steel wire about yay big.
- 20:11And it gets used to do just about everything.
- 20:13Its original purpose was for fencing.
- 20:15But yeah, we, there's just just a million different applications
- 20:19for it globally. But yeah, we, we sort of built
- 20:23ourselves to be a nation of #8 wire thinking.
- 20:26Yeah. Super like super practical.
- 20:30Yeah, practical outside of the box.
- 20:33I think something else as well we see down here is a lot of
- 20:37purpose driven, like a lot of purpose driven work.
- 20:40And so we not everything is just about growing for the sake of
- 20:43growing. A lot of us are very mission
- 20:45driven. We do things because we want to
- 20:48do them because they make the world a better place.
- 20:51And and as a result of that, there are like some truly
- 20:54outstanding companies sort of coming out of New Zealand.
- 20:58And most of those, a lot of the sort of the series being above
- 21:05have ended up coming on board with us because of the growths
- 21:07that they're going through internationally.
- 21:09Yeah. And it's just, it's a really
- 21:11different sort of way of thinking and working.
- 21:13Moral of the story? If you're hesitating to be an
- 21:17entrepreneur in the US, move to New Zealand.
- 21:19You'll have lots of opportunity there.
- 21:23Yeah, for sure. Next question is, is I learned
- 21:27about a new phrase called tall poppy syndrome.
- 21:31And for all listeners, that is where you know, people who stand
- 21:37out for their success ambition are criticized or suddenly cut
- 21:41down kind of like tall puppies in the field.
- 21:43The the field gets mowed so that everybody gets is even.
- 21:47Tell me about whether that has made you more cautious or
- 21:52grounded, or if that worries you at all.
- 21:55I think it used to be a real problem in New Zealand as of
- 21:59late. Our tech ecosystem has really
- 22:02sort of to right itself and there's a lot of support.
- 22:06Everybody's sort of willing to put a hand up and do what they
- 22:08can to help. And I think that was sort of
- 22:11just a result of the immaturity, one, from the business and two,
- 22:16from an ecosystem standpoint in this country in terms of how it
- 22:19shaped me. That's a good question.
- 22:23I don't think it's really ever bothered me too much.
- 22:25I, I care really deeply about the problem I'm solving.
- 22:30I'm happened to be working on a really big problem.
- 22:33It's, you know, because the world, $5 trillion a year wasted
- 22:37spend on software, wasted opportunity.
- 22:41And that's all I've really ever worried about is working on
- 22:43things that I care about. If I am doing that, Yeah, I
- 22:47don't know. It just it's never really
- 22:48occurred to me what other people think about it.
- 22:51Yeah, that's great. So you kind of have your own,
- 22:54you know, your own confidence and self assurance about about
- 22:56what you're working. On I think so long as I am able
- 22:59to solve the problem well, that's all I've really cared
- 23:02about in founding this business. The thing we've built everything
- 23:06on top of was being able to do what we say we can do.
- 23:10And so very much like a customer success first mentality, which
- 23:14was I guess where my career started.
- 23:17And yeah, it's sort of yielding itself and the results that we
- 23:19average out 1200% ROI in the first year with our customers.
- 23:25We've net net revenue retention of 147 percent, 100% logo
- 23:30retention. And I think that's just sort of
- 23:32down to the way we think about how we approach things, which I
- 23:35would say is also very key. You know, it's like we like to
- 23:38sort of think about trust, do good business.
- 23:44OK, lightning round, flat, white or long blacks.
- 23:48Long black. That was cream.
- 23:51Black with cream. OK, what about jandals or
- 23:54trainers? Jandals, definitely.
- 23:57At the moment, it's summer. OK.
- 23:59And all black or black caps. All Blacks, yeah.
- 24:07OK. Yeah.
- 24:08Last question here. To rack up when you travel and
- 24:12you come home to New Zealand, what's the first thing that you
- 24:18do to reset yourself? Go to the beach.
- 24:22Yeah, we. I'm surrounded.
- 24:23I'm ugly enough to have I think 9 or 10 beaches within about 10
- 24:27minutes drive of my house. So yeah, it's definitely where
- 24:32where the heart is. We're very lucky.
- 24:35New Zealand's an island nation, so they're just if you drive for
- 24:37an hour in any direction, you'll probably end up at a beach.
- 24:41Awesome. Steve, thank you so much for
- 24:43being with us. Is there anything else that
- 24:44you'd like to tell the audience before he calls out here?
- 24:48No, I think that's it. Really enjoyed the conversation
- 24:50today. Laura, thank you so much for
- 24:51your time of. Course, thanks for being on the
- 24:53show to all listeners. If you enjoyed this episode,
- 24:57please give us a like or comment and so that we know we liked it
- 25:00and we can give you more content.
- 25:02Otherwise, happy New Year and we will see you next time.
- 25:05Thank you so much.