AI in Product Development: Enhancing Workflows & Minimum Iteration w/ Cody Schneider

[00:00:02] [SPEAKER_01]: Let's do a little bit more AI.

[00:00:04] [SPEAKER_01]: Let's talk to a developer who is leading a couple of product companies who have been building

[00:00:09] [SPEAKER_01]: different solutions.

[00:00:11] [SPEAKER_01]: What are his prerequisites?

[00:00:12] [SPEAKER_01]: How is he measuring success?

[00:00:15] [SPEAKER_01]: And what is necessary to build one that's successful?

[00:00:18] [SPEAKER_01]: Cody Schneider joins me today on this bonus episode of the Business of Tech.

[00:00:25] [SPEAKER_01]: Today's episode is supported by CoreView.

[00:00:27] [SPEAKER_01]: Your customers need your Microsoft 365 expertise and CoreView has the only M365 management platform

[00:00:35] [SPEAKER_01]: designed for MSPs.

[00:00:37] [SPEAKER_01]: Manage hundreds of tenants, automate manual tasks, and monitor compliance all while intelligently

[00:00:43] [SPEAKER_01]: comparing to the baseline.

[00:00:44] [SPEAKER_01]: With a no-code control approach, CoreView revolutionizes your Microsoft 365 administration.

[00:00:50] [SPEAKER_01]: This powerful platform enables automatic reporting and remediation, ensuring optimal performance

[00:00:56] [SPEAKER_01]: and security.

[00:00:57] [SPEAKER_01]: The best part?

[00:00:59] [SPEAKER_01]: You achieve this high level of service without the need for a large workforce, allowing

[00:01:03] [SPEAKER_01]: you to focus on growing your business through efficiency.

[00:01:07] [SPEAKER_01]: Want to know more?

[00:01:08] [SPEAKER_01]: Visit coreview.com slash MSP and find out more.

[00:01:15] [SPEAKER_01]: Cody, well thanks for joining me today.

[00:01:17] [SPEAKER_01]: Thanks for having me Dave.

[00:01:18] [SPEAKER_01]: Super excited to be here.

[00:01:20] [SPEAKER_01]: So I'm going to dive sort of right in.

[00:01:22] [SPEAKER_01]: You've been thinking a lot about the way you can automate and enhance workflows using artificial

[00:01:29] [SPEAKER_01]: intelligence.

[00:01:30] [SPEAKER_01]: Tell me a little bit about what you're doing there and what your framework kind of for success

[00:01:34] [SPEAKER_01]: and what you're looking for in the workflow space is.

[00:01:38] [SPEAKER_00]: Yeah, absolutely.

[00:01:40] [SPEAKER_00]: So I think that the high level of all this is that AI is incredible at taking unstructured

[00:01:45] [SPEAKER_00]: data and structuring it.

[00:01:46] [SPEAKER_00]: So how I see the best implementations occurring currently is you look at a current processor

[00:01:52] [SPEAKER_00]: system that an organization has, you audit that, you take a step back and you're like,

[00:01:56] [SPEAKER_00]: okay, what tools can I implement within this workflow to try to automate it as much of

[00:02:00] [SPEAKER_00]: this as way?

[00:02:01] [SPEAKER_00]: And then this really enables the humans that are involved in that process to focus

[00:02:06] [SPEAKER_00]: on those human components.

[00:02:07] [SPEAKER_00]: So you can example real estate and sales, right?

[00:02:11] [SPEAKER_00]: A lot of paperwork that's typically happening on the back end of that.

[00:02:13] [SPEAKER_00]: AI is a great solution for that paperwork component of, oh, I have all this information.

[00:02:18] [SPEAKER_00]: I'm trying to get it into the structured PDF format as an example.

[00:02:22] [SPEAKER_00]: And then that allows for that sales director or whoever is on the front lines doing those

[00:02:27] [SPEAKER_00]: calls to focus on those high touch human parts of this business.

[00:02:31] [SPEAKER_00]: And that's how we can cut costs and also improve that efficiency within those companies.

[00:02:36] [SPEAKER_00]: So I think that's the biggest thing though is like that unstructured data is structured.

[00:02:40] [SPEAKER_00]: That is what we're seeing be the most impactful things that these companies can do to begin with.

[00:02:46] [SPEAKER_01]: Now, when we talk about unstructured data, is it just chaotic, massive data?

[00:02:52] [SPEAKER_01]: 100%

[00:02:53] [SPEAKER_01]: Throw it in there or but there has to be some prep, right?

[00:02:56] [SPEAKER_01]: In order to get things to be useful, we have to at least take a look at it and

[00:02:59] [SPEAKER_01]: understand like, well, we don't want to give everyone access to everything.

[00:03:02] [SPEAKER_01]: We perhaps want to think about access controls.

[00:03:05] [SPEAKER_01]: Like talk to me a little bit about the process of data management that you think

[00:03:09] [SPEAKER_01]: has to happen to be successful doing something with AI.

[00:03:14] [SPEAKER_00]: Yeah, I think on the data management side, of course, certain documents within

[00:03:20] [SPEAKER_00]: the organization need to have the same access controls, etc.

[00:03:24] [SPEAKER_00]: I think the bigger thing that again, we work a lot in the content space, right?

[00:03:29] [SPEAKER_00]: So a lot of our source data or source files that we're touching is

[00:03:33] [SPEAKER_00]: a lot of its public information like it's a podcast you could go and download it

[00:03:37] [SPEAKER_00]: and transform it.

[00:03:39] [SPEAKER_00]: But the bigger thing is that we can give all of this context.

[00:03:44] [SPEAKER_00]: So as an example, say you wanted to write a blog post, right?

[00:03:47] [SPEAKER_00]: And I go and I scrape the front page of Google for a target keyword phrase.

[00:03:51] [SPEAKER_00]: So I get the transcripts from the YouTube videos.

[00:03:53] [SPEAKER_00]: I get all the blog post content from that.

[00:03:55] [SPEAKER_00]: I put that in the context window and then I say, okay,

[00:03:57] [SPEAKER_00]: write me a detailed blog post outline for this target keyword,

[00:04:02] [SPEAKER_00]: writes the outline and then I can prompt it now, write the article.

[00:04:05] [SPEAKER_00]: I don't have to structure that data perfectly on that basically first go,

[00:04:11] [SPEAKER_00]: whatever in that context window to extract that knowledge.

[00:04:14] [SPEAKER_00]: And I think that's the piece where I'm seeing companies start to realize this

[00:04:18] [SPEAKER_00]: and it's like, oh, I have all these PDFs.

[00:04:21] [SPEAKER_00]: Maybe it's internal documentation on processes or white papers that we wrote.

[00:04:26] [SPEAKER_00]: I can take these 10 different white papers based off of this current client situation.

[00:04:29] [SPEAKER_00]: I can now create a custom report or a custom output for them by providing this

[00:04:36] [SPEAKER_00]: unstructured version of this.

[00:04:38] [SPEAKER_00]: It doesn't have to be perfect.

[00:04:39] [SPEAKER_00]: It can extract that knowledge and information and then transform it or

[00:04:42] [SPEAKER_00]: translate it into the format that you're trying to get it.

[00:04:45] [SPEAKER_00]: And I think how, especially when you kind of scale this up to a more technical part,

[00:04:50] [SPEAKER_00]: like even providing like, here's a JSON format, right?

[00:04:53] [SPEAKER_00]: That I want you to output almost that scaffolding.

[00:04:55] [SPEAKER_00]: Like here's the skeleton of what I want this document that you're going to generate

[00:04:59] [SPEAKER_00]: to look like and then providing that source material that it can then

[00:05:03] [SPEAKER_00]: extract and pull from it.

[00:05:05] [SPEAKER_00]: That's how, again, we're seeing these organizations basically take all of this

[00:05:11] [SPEAKER_00]: disparate information and then consolidate it into a new piece of content

[00:05:17] [SPEAKER_00]: or a new, whatever that ends up being within the organization.

[00:05:21] [SPEAKER_01]: Now, one of the things that you mentioned some of the interesting sort of sales

[00:05:26] [SPEAKER_01]: scenarios, the workflow scenarios.

[00:05:28] [SPEAKER_01]: And we oftentimes are hearing a lot of the sort of hype ones too.

[00:05:31] [SPEAKER_01]: You also talked about like the content transformation ones.

[00:05:34] [SPEAKER_01]: How are you differentiating between sort of that hype and practical applications

[00:05:39] [SPEAKER_01]: for AI?

[00:05:40] [SPEAKER_01]: Like where are the areas that you find most useful and what are your criteria for that?

[00:05:44] [SPEAKER_01]: Absolutely.

[00:05:45] [SPEAKER_00]: The most impactful places that this can be.

[00:05:48] [SPEAKER_00]: So like if I just go and I prompt the AI to do something, right?

[00:05:50] [SPEAKER_00]: It's writing from its general knowledge.

[00:05:52] [SPEAKER_00]: The general knowledge of the internet is very average where all of the value is

[00:05:57] [SPEAKER_00]: in the tails, right?

[00:05:58] [SPEAKER_00]: So expert knowledge has actually become more valuable than ever before.

[00:06:03] [SPEAKER_00]: And so what we see is the real value is when you have your own data sets,

[00:06:07] [SPEAKER_00]: you have your own libraries of content,

[00:06:09] [SPEAKER_00]: you have your own proprietary information that you're then using

[00:06:12] [SPEAKER_00]: AI to work off of, right?

[00:06:14] [SPEAKER_00]: It's using its general knowledge to generate some type of new thing from

[00:06:18] [SPEAKER_00]: that source material or files that you provide it.

[00:06:23] [SPEAKER_00]: That is actually where all the value comes from.

[00:06:27] [SPEAKER_00]: Like there's a, you know, we see this right now in the sales component

[00:06:30] [SPEAKER_00]: like where it's like, hey, we're going to automate an BDR role or an SDR role

[00:06:33] [SPEAKER_00]: entirely.

[00:06:34] [SPEAKER_00]: Like you're going to not even have to employ this person.

[00:06:36] [SPEAKER_00]: It's like actually probably not maybe five years from now,

[00:06:40] [SPEAKER_00]: but what's way more effective is like what's a task that this person

[00:06:43] [SPEAKER_00]: is doing currently that they hate doing.

[00:06:46] [SPEAKER_00]: Like a lot of times you can just go into the organization and be like, hey,

[00:06:49] [SPEAKER_00]: you know, Jennifer, like what do you hate doing on a daily basis?

[00:06:52] [SPEAKER_00]: She's like, well, this is the massive time suck and it's mind numbing work.

[00:06:56] [SPEAKER_00]: You know, nine times out of 10, we see that like there's some way to

[00:06:58] [SPEAKER_00]: typically like automate or augment that workflow that they're currently

[00:07:03] [SPEAKER_00]: creating.

[00:07:04] [SPEAKER_00]: And that is where it's the most impactful.

[00:07:07] [SPEAKER_00]: Like that is especially on a team management side and make clients

[00:07:10] [SPEAKER_00]: successful and happy.

[00:07:12] [SPEAKER_00]: And I mean, employee turnover is like a huge problem right in the U.S.

[00:07:16] [SPEAKER_00]: just looking within the workforce.

[00:07:17] [SPEAKER_00]: If we can get rid of these things that they despise about their day job

[00:07:21] [SPEAKER_00]: or going to work, that's a way for us to increase that retention reduces

[00:07:26] [SPEAKER_00]: the costs on the HR department for basically having those teams leave.

[00:07:31] [SPEAKER_00]: And so I think that's again, the things that we're seeing on the

[00:07:34] [SPEAKER_00]: front line that companies are implementing that are they're

[00:07:38] [SPEAKER_00]: seeing success with and affecting their bottom line of positively.

[00:07:43] [SPEAKER_01]: So it's interesting because you said in the in there, you know, you talked

[00:07:46] [SPEAKER_01]: about how knowledge is the most valuable stuff and I would agree with

[00:07:50] [SPEAKER_01]: you by the way, I think expertise is where is bubbling up is even

[00:07:52] [SPEAKER_01]: more important.

[00:07:54] [SPEAKER_01]: And you also said correctly, and I would agree with you that like

[00:07:56] [SPEAKER_01]: the average AI output is intensely average.

[00:08:01] [SPEAKER_01]: Right?

[00:08:01] [SPEAKER_01]: So how do you view kind of bringing those two things together if

[00:08:05] [SPEAKER_01]: the AI's output is intensely average?

[00:08:09] [SPEAKER_01]: But at the same time, we're trying to make, you know, people

[00:08:12] [SPEAKER_01]: and knowledge more valuable.

[00:08:14] [SPEAKER_01]: How do you see those two working together best?

[00:08:18] [SPEAKER_00]: For sure.

[00:08:19] [SPEAKER_00]: I think the expert level of this is you're training your own models,

[00:08:22] [SPEAKER_00]: right?

[00:08:23] [SPEAKER_00]: Like you're basically modifying the base models in some way for

[00:08:26] [SPEAKER_00]: your companies.

[00:08:27] [SPEAKER_00]: Striper is a great example of this where they basically took chat

[00:08:30] [SPEAKER_00]: GPT and then they fine tuned it so that it was, you know,

[00:08:33] [SPEAKER_00]: the chat bot that you're working with internally is it feels

[00:08:38] [SPEAKER_00]: on brand.

[00:08:39] [SPEAKER_00]: That makes sense.

[00:08:41] [SPEAKER_00]: Aeroelectronics is another huge company, you know, they resell

[00:08:43] [SPEAKER_00]: components, same idea.

[00:08:45] [SPEAKER_00]: They have like an internal tool that they built.

[00:08:46] [SPEAKER_00]: It's just basically a fine tuned version of that.

[00:08:48] [SPEAKER_00]: The cheating way to do this, you don't want to go fine tune

[00:08:51] [SPEAKER_00]: it because it does take training data and just, you know,

[00:08:54] [SPEAKER_00]: basically energy, right?

[00:08:56] [SPEAKER_00]: Like which is one of the most valued commodities at this

[00:09:02] [SPEAKER_00]: point.

[00:09:03] [SPEAKER_00]: Everybody's fighting for all this GPU power, but the other

[00:09:08] [SPEAKER_00]: way around this is you can basically like, for example,

[00:09:11] [SPEAKER_00]: say I provide this transcript and it has all this unique

[00:09:13] [SPEAKER_00]: vocabulary within it.

[00:09:14] [SPEAKER_00]: I can say, hey, you know, use vocabulary from that source

[00:09:17] [SPEAKER_00]: material with what you're writing.

[00:09:20] [SPEAKER_00]: Also here's your persona.

[00:09:21] [SPEAKER_00]: So like defining who that, like what the person is that

[00:09:24] [SPEAKER_00]: you're writing from that angle.

[00:09:26] [SPEAKER_00]: And then the other piece is then you can even like

[00:09:30] [SPEAKER_00]: define the tone, style and voice based off of, okay.

[00:09:34] [SPEAKER_00]: The way I try to talk about it is like traditionally you

[00:09:37] [SPEAKER_00]: would go make like a brand style guide as an example

[00:09:39] [SPEAKER_00]: within an organization.

[00:09:41] [SPEAKER_00]: You're now going to have like an AI style guide, same

[00:09:44] [SPEAKER_00]: concept, but it's basically these base level prompts

[00:09:47] [SPEAKER_00]: that your whole team is interacting with these tools

[00:09:50] [SPEAKER_00]: with and that's how you create consistency.

[00:09:52] [SPEAKER_00]: That's how you can create better outputs.

[00:09:54] [SPEAKER_00]: And again, it's like everything and, you know,

[00:09:57] [SPEAKER_00]: anything in technology we work with, especially on

[00:09:59] [SPEAKER_00]: the data science side, it's garbage in, garbage out.

[00:10:01] [SPEAKER_00]: Like content quality is super, whatever you provide it

[00:10:05] [SPEAKER_00]: is super important to create like good outputs.

[00:10:08] [SPEAKER_00]: But once you have in that library, that back

[00:10:13] [SPEAKER_00]: catalog information or knowledge, it's very capable

[00:10:16] [SPEAKER_00]: at understanding, oh, like here's the language

[00:10:19] [SPEAKER_00]: that it speaks in.

[00:10:20] [SPEAKER_00]: Here's the context for me to, for whatever it is

[00:10:24] [SPEAKER_00]: I'm working on or producing from an output standpoint

[00:10:26] [SPEAKER_00]: and you're just going to get a way higher quality piece

[00:10:30] [SPEAKER_00]: of generated whatever it creates when you provide

[00:10:36] [SPEAKER_00]: that source material that you have, that is to your

[00:10:39] [SPEAKER_00]: standard, that is to your quality level.

[00:10:41] [SPEAKER_00]: And to make a cron-created example of this,

[00:10:43] [SPEAKER_00]: like say you, SEO was a huge thing that your company

[00:10:46] [SPEAKER_00]: was doing, you were writing a ton of blog posts

[00:10:48] [SPEAKER_00]: that was like where you're getting inbound traffic.

[00:10:50] [SPEAKER_00]: You can go and show the AI, hey, here's all these

[00:10:53] [SPEAKER_00]: blog posts, we'll say our top 20 best performing

[00:10:57] [SPEAKER_00]: create a template outline that I can use in the future

[00:11:01] [SPEAKER_00]: for how blog posts should be structured.

[00:11:03] [SPEAKER_00]: Okay, cool, now I want to go write another one.

[00:11:05] [SPEAKER_00]: I take that template and outline and I take the language

[00:11:08] [SPEAKER_00]: and the tone and the style and the voice

[00:11:10] [SPEAKER_00]: and I apply that all together into that composite

[00:11:13] [SPEAKER_00]: and what you're going to get from that.

[00:11:15] [SPEAKER_00]: I mean it's going to be in the top 10%

[00:11:18] [SPEAKER_00]: from a content standpoint in comparison to if I just

[00:11:20] [SPEAKER_00]: tried to one shot single prompt the AI

[00:11:23] [SPEAKER_00]: to write a document for me, right?

[00:11:25] [SPEAKER_00]: It has context, it has more understanding

[00:11:28] [SPEAKER_00]: of what the expectations and the outputs are

[00:11:30] [SPEAKER_00]: and I almost try to get people to think about it

[00:11:32] [SPEAKER_00]: like hey, it's a junior team member, right?

[00:11:34] [SPEAKER_00]: Because as detailed and step by step as you can make it,

[00:11:38] [SPEAKER_00]: that's how you're going to get the best outcomes

[00:11:40] [SPEAKER_00]: from a generation standpoint.

[00:11:42] [SPEAKER_01]: Interestingly, I'm going to shift gears a little bit

[00:11:45] [SPEAKER_01]: because you know, and I'm going to have to give

[00:11:46] [SPEAKER_01]: some listeners some a little bit of exposure

[00:11:48] [SPEAKER_01]: to the way this works.

[00:11:49] [SPEAKER_01]: Actually, I get pitched a lot of ideas of topics from teams

[00:11:53] [SPEAKER_01]: and your team reached out with a really interesting one

[00:11:55] [SPEAKER_01]: and as I was prepping to get this ready

[00:11:57] [SPEAKER_01]: I actually have to disclose I realized

[00:11:59] [SPEAKER_01]: I was a customer of yours because one of the products

[00:12:01] [SPEAKER_01]: you work on, Swell AI is something that I use in my workflow.

[00:12:05] [SPEAKER_01]: The reason that I wanted to talk to you was

[00:12:07] [SPEAKER_01]: the approach you do to product development.

[00:12:10] [SPEAKER_01]: So you've got a couple of different products

[00:12:11] [SPEAKER_01]: that you work on and you have a theme

[00:12:14] [SPEAKER_01]: of minimum iteration and I was really intrigued

[00:12:17] [SPEAKER_01]: about this idea and it's one of the reasons

[00:12:20] [SPEAKER_01]: I wanted to have you on the show.

[00:12:21] [SPEAKER_01]: Tell me more about this idea of minimum iteration.

[00:12:25] [SPEAKER_00]: Yeah, so it's like every early stage company,

[00:12:29] [SPEAKER_00]: we just try to get things in public as fast as possible.

[00:12:35] [SPEAKER_00]: I think very often and I'm a problem of this as well.

[00:12:40] [SPEAKER_00]: It's like I know where this is going and that's

[00:12:44] [SPEAKER_00]: that North Star goal but it's like what are these small steps

[00:12:47] [SPEAKER_00]: to get there and in reality,

[00:12:50] [SPEAKER_00]: you think you're heading north but in reality

[00:12:52] [SPEAKER_00]: you need to head northwest.

[00:12:53] [SPEAKER_00]: You're still going north but it's almost an entirely

[00:12:56] [SPEAKER_00]: different direction.

[00:12:56] [SPEAKER_00]: You're going to end up in a different place

[00:12:57] [SPEAKER_00]: but you don't know that until you take a step forward.

[00:13:00] [SPEAKER_00]: And so relating that to the product development side

[00:13:03] [SPEAKER_00]: that's how we try to approach all of this

[00:13:04] [SPEAKER_00]: is basically as fast as we can.

[00:13:07] [SPEAKER_00]: How do we get it in the hands of the people

[00:13:08] [SPEAKER_00]: that are using it and we get that feedback

[00:13:11] [SPEAKER_00]: from them.

[00:13:15] [SPEAKER_00]: It's the only way that we found to actually make things

[00:13:18] [SPEAKER_00]: that people want to buy.

[00:13:19] [SPEAKER_00]: We have theories about what people want to purchase

[00:13:23] [SPEAKER_00]: but in reality until it's in the public

[00:13:25] [SPEAKER_00]: and they can interact with it, you have no idea.

[00:13:27] [SPEAKER_00]: And so that's how we try to lean into that.

[00:13:30] [SPEAKER_00]: I hope that's why you met on the minimum iteration side.

[00:13:33] [SPEAKER_00]: I don't know if I took that in a completely

[00:13:34] [SPEAKER_00]: incorrect direction but...

[00:13:37] [SPEAKER_01]: No it is and what I wanted to sort of say

[00:13:41] [SPEAKER_01]: is there some best practice that you've learned

[00:13:44] [SPEAKER_01]: in doing this process that you can say like

[00:13:47] [SPEAKER_01]: this is what you should think about.

[00:13:49] [SPEAKER_01]: It's similar to minimum viable product

[00:13:52] [SPEAKER_01]: and it's an approach to development.

[00:13:54] [SPEAKER_01]: What's the best practice that you would tell

[00:13:56] [SPEAKER_01]: to those that are thinking about this

[00:13:58] [SPEAKER_01]: particularly as they apply it to services

[00:13:59] [SPEAKER_01]: that you've learned over the time using this approach?

[00:14:03] [SPEAKER_00]: Yeah, that's a great question.

[00:14:05] [SPEAKER_00]: I think for me how I approach this

[00:14:09] [SPEAKER_00]: and think about this is you're just trying

[00:14:12] [SPEAKER_00]: to get it good enough.

[00:14:13] [SPEAKER_00]: You can always improve that, right?

[00:14:15] [SPEAKER_00]: But if I can get it 80% of the way there

[00:14:16] [SPEAKER_00]: in 24 hours and have that live

[00:14:19] [SPEAKER_00]: and in the hands of people,

[00:14:22] [SPEAKER_00]: that's going to be so much more impactful

[00:14:24] [SPEAKER_00]: for my business than it's perfect a month from now.

[00:14:27] [SPEAKER_00]: We're going to build this thing

[00:14:28] [SPEAKER_00]: and then we're probably going to have to take steps back

[00:14:30] [SPEAKER_00]: because we've overbuilt it

[00:14:32] [SPEAKER_00]: and then that just creates more dev time in the aggregate.

[00:14:36] [SPEAKER_00]: But if I just put it in the hands of these people

[00:14:37] [SPEAKER_00]: as fast as I could and then allow them to kind of finish

[00:14:40] [SPEAKER_00]: tie up the loose ends on that live city percent.

[00:14:43] [SPEAKER_00]: This all depends as well on the company size

[00:14:46] [SPEAKER_00]: and like, yeah, I've done work in,

[00:14:50] [SPEAKER_00]: for orbs that are S&P 500,

[00:14:53] [SPEAKER_00]: like it's different rules

[00:14:55] [SPEAKER_00]: in the pace of all this change

[00:14:56] [SPEAKER_00]: but if you're working with small businesses, right?

[00:14:58] [SPEAKER_00]: They're way more flexible.

[00:14:59] [SPEAKER_00]: They're way more receptive to being like,

[00:15:00] [SPEAKER_00]: oh, this is good enough.

[00:15:02] [SPEAKER_00]: And then we're going to iron out all of the wrinkles

[00:15:08] [SPEAKER_00]: as we get this into public.

[00:15:10] [SPEAKER_00]: But I think that especially on client work

[00:15:12] [SPEAKER_00]: and this is just what I found being in this world for so long,

[00:15:15] [SPEAKER_00]: speed of action is like the biggest measurement

[00:15:18] [SPEAKER_00]: of client satisfaction.

[00:15:20] [SPEAKER_00]: Like if they ask for something

[00:15:21] [SPEAKER_00]: or you promise something,

[00:15:22] [SPEAKER_00]: like how quickly you can get that out to them

[00:15:24] [SPEAKER_00]: is what's...

[00:15:26] [SPEAKER_00]: That's like the only thing that I've ever found

[00:15:28] [SPEAKER_00]: to be directly correlated with reducing churn

[00:15:31] [SPEAKER_00]: and the longevity of that client staying with you, right?

[00:15:34] [SPEAKER_00]: And so when you think about that

[00:15:35] [SPEAKER_00]: in relationships with products

[00:15:36] [SPEAKER_00]: or implementing these services for your clients,

[00:15:39] [SPEAKER_00]: it's then coming back to, okay,

[00:15:41] [SPEAKER_00]: well what is the...

[00:15:42] [SPEAKER_00]: Even if it's not all the way they're perfect

[00:15:44] [SPEAKER_00]: and this is hard for everybody, right?

[00:15:46] [SPEAKER_00]: It's that we want to do good work

[00:15:49] [SPEAKER_00]: and we want it to be...

[00:15:53] [SPEAKER_00]: That's...

[00:15:53] [SPEAKER_00]: Our work is basically a measurement of our value,

[00:15:56] [SPEAKER_00]: like the quality that we put out there.

[00:15:58] [SPEAKER_00]: But I think that with...

[00:16:00] [SPEAKER_00]: Especially with this newer technology,

[00:16:01] [SPEAKER_00]: there's no way to make it perfect.

[00:16:02] [SPEAKER_00]: Like we've even found this where it's like,

[00:16:05] [SPEAKER_00]: you know, every once in a while

[00:16:06] [SPEAKER_00]: it's just gonna totally mess up, right?

[00:16:08] [SPEAKER_00]: And we just have to kind of be okay with that.

[00:16:10] [SPEAKER_00]: But again, if it's 80% of the weight there,

[00:16:14] [SPEAKER_00]: especially for these companies where it's like,

[00:16:16] [SPEAKER_00]: oh, I just automated a salary away, right?

[00:16:18] [SPEAKER_00]: Like, of course they're gonna be ecstatic

[00:16:21] [SPEAKER_00]: and also that creates just a high impact activity

[00:16:24] [SPEAKER_00]: that you can do for these people.

[00:16:25] [SPEAKER_00]: So I think the biggest thing is that it's just like fighting

[00:16:29] [SPEAKER_00]: and so many people are focusing on that perfection.

[00:16:33] [SPEAKER_00]: But in reality, like building good product is iteration.

[00:16:36] [SPEAKER_00]: Like it's that process of how many revolutions might...

[00:16:39] [SPEAKER_00]: My friend's in the wine industry actually

[00:16:40] [SPEAKER_00]: and he told this to me like two years ago.

[00:16:44] [SPEAKER_00]: He was like, hey, I potentially get to go and make wine

[00:16:48] [SPEAKER_00]: 40 more times in my life, right?

[00:16:51] [SPEAKER_00]: Like they see it once a year.

[00:16:52] [SPEAKER_00]: And so for him, he's constantly talking to people

[00:16:55] [SPEAKER_00]: to try to get as much information from other winemakers

[00:16:59] [SPEAKER_00]: so that he can have more at bats is how he described it.

[00:17:04] [SPEAKER_00]: I've like seeing more pitches, right?

[00:17:06] [SPEAKER_00]: Because the more pitches you see,

[00:17:07] [SPEAKER_00]: like the more likely you are to hit a home run

[00:17:10] [SPEAKER_00]: or hit a bit, you know, get on base basically.

[00:17:12] [SPEAKER_00]: And I think it's that same idea when you think about iteration.

[00:17:15] [SPEAKER_00]: You're just trying to see as many cycles as possible

[00:17:17] [SPEAKER_00]: because the faster the cycles can be,

[00:17:20] [SPEAKER_00]: the better the outcomes are going to be.

[00:17:22] [SPEAKER_00]: It's also going to feel like movement is happening for the client

[00:17:25] [SPEAKER_00]: and they see that progress, they see that progression

[00:17:28] [SPEAKER_00]: and that excitement creates stickiness from my experience

[00:17:31] [SPEAKER_00]: within the client experience,

[00:17:33] [SPEAKER_00]: whether it's a product I service

[00:17:35] [SPEAKER_00]: or you know, a software that you're selling them, etc.

[00:17:37] [SPEAKER_00]: I think it's the same thing.

[00:17:38] [SPEAKER_00]: People get... They want to be a part of something, right?

[00:17:40] [SPEAKER_00]: And when you... When they see that movement,

[00:17:42] [SPEAKER_00]: they see that momentum.

[00:17:43] [SPEAKER_00]: It feels like they're a part of something

[00:17:45] [SPEAKER_00]: that's larger than themselves.

[00:17:46] [SPEAKER_00]: So that's a way to lean into it.

[00:17:47] [SPEAKER_00]: A lot of psychological stuff in there

[00:17:49] [SPEAKER_00]: but again, just brain dumping all the things

[00:17:51] [SPEAKER_00]: that I found to be relevant and related kind of

[00:17:53] [SPEAKER_00]: from during my career.

[00:17:55] So...

[00:17:56] [SPEAKER_01]: Well, you've worked with a lot of clients

[00:17:57] [SPEAKER_01]: and you've talked about the different approaches.

[00:17:59] [SPEAKER_01]: I want to ask then about like that you mentioned the...

[00:18:02] [SPEAKER_01]: You have to be comfortable with it

[00:18:03] [SPEAKER_01]: with these kind of products being wrong.

[00:18:05] [SPEAKER_01]: Now for a lot of people that are very used

[00:18:07] [SPEAKER_01]: to the technology being that very...

[00:18:10] [SPEAKER_01]: You know, that very regimented.

[00:18:13] [SPEAKER_01]: You put it in input,

[00:18:14] [SPEAKER_01]: you will get the same output every single time.

[00:18:16] [SPEAKER_01]: You know, this is very different than that.

[00:18:19] [SPEAKER_01]: How are you finding it best to make customers

[00:18:23] [SPEAKER_01]: and clients comfortable with that level

[00:18:27] [SPEAKER_01]: of potential chaos or the...

[00:18:29] [SPEAKER_01]: The introduction of the...

[00:18:31] [SPEAKER_01]: Is now less reliable in order to get to achieving it?

[00:18:36] [SPEAKER_01]: Like how are you introducing and wrestling with that?

[00:18:39] [SPEAKER_00]: Yeah.

[00:18:39] [SPEAKER_00]: So I'm positioning it in people's minds as like,

[00:18:41] [SPEAKER_00]: hey, your team is now a curator more than anything, right?

[00:18:45] [SPEAKER_00]: Rather than a producer, like you're going to lean on these tools

[00:18:48] [SPEAKER_00]: to do all the grunt labor heavy lifting, right?

[00:18:51] [SPEAKER_00]: So again, if maybe it's data entry,

[00:18:52] [SPEAKER_00]: maybe it's extraction of information from a huge dataset,

[00:18:57] [SPEAKER_00]: whatever that ends up being,

[00:18:59] [SPEAKER_00]: your team is now going in

[00:19:00] [SPEAKER_00]: and they're doing the verification of these things

[00:19:02] [SPEAKER_00]: to confirm that, oh, this looks like...

[00:19:05] [SPEAKER_00]: Right, this looks real, right?

[00:19:06] [SPEAKER_00]: And for...

[00:19:07] [SPEAKER_00]: Like what we see is like a lot of the time

[00:19:09] [SPEAKER_00]: and money that spends in most organizations

[00:19:11] [SPEAKER_00]: is on that kind of grunt labor component of this.

[00:19:14] [SPEAKER_00]: But if I can get that to where it's 90% of the way there

[00:19:17] [SPEAKER_00]: and then I have a 10x employee come in

[00:19:19] [SPEAKER_00]: and they get that hand off for that piece of whatever it is

[00:19:23] [SPEAKER_00]: that I'm trying to produce,

[00:19:24] [SPEAKER_00]: what we're going to see is like we can make that 10x employee

[00:19:27] [SPEAKER_00]: a 100x employee and suddenly we've just made it a company

[00:19:31] [SPEAKER_00]: where it's like say you have a headcount of 50 people.

[00:19:33] [SPEAKER_00]: Well, this organization is now functioning

[00:19:35] [SPEAKER_00]: like a 500 person organization.

[00:19:37] [SPEAKER_00]: What does that mean to the revenues?

[00:19:38] [SPEAKER_00]: What does that mean to the bottom line of these organizations?

[00:19:40] [SPEAKER_00]: And I think that framing of like your...

[00:19:45] [SPEAKER_00]: The AI is not perfect, right?

[00:19:47] [SPEAKER_00]: But it can get good enough, like get things to a place

[00:19:51] [SPEAKER_00]: where they're good enough that a human can then take that hand off

[00:19:54] [SPEAKER_00]: and they can put all of the magic into that, right?

[00:19:56] [SPEAKER_00]: And we see this with content that it's producing, reports or whatever.

[00:20:00] [SPEAKER_00]: There's these human...

[00:20:01] [SPEAKER_00]: It's like flair, right?

[00:20:03] [SPEAKER_00]: That the AI is really...

[00:20:05] [SPEAKER_00]: It's very hard to get it to do it consistently.

[00:20:07] [SPEAKER_00]: We can get it to do kind of these rote,

[00:20:11] [SPEAKER_00]: almost militant tasks very effectively and they never complain.

[00:20:16] [SPEAKER_00]: And they don't write a negative review about your company on Glassdoor.

[00:20:20] [SPEAKER_00]: And so that is...

[00:20:21] [SPEAKER_00]: I think the framing that I'm trying to brainwash

[00:20:25] [SPEAKER_00]: as many founders to think about is how do I make my team

[00:20:30] [SPEAKER_00]: that's already exceptional,

[00:20:32] [SPEAKER_00]: how do I make them perform like they're almost superhuman

[00:20:34] [SPEAKER_00]: with augmenting them with these tools?

[00:20:37] [SPEAKER_01]: Well, Cody Snider is the CEO of Draft Horse AI and Swell AI.

[00:20:42] [SPEAKER_01]: He's known for his innovative approach to rapidly scaling startups

[00:20:45] [SPEAKER_01]: and redefining digital processes through AI.

[00:20:49] [SPEAKER_01]: His focus is on augmenting workflows and boosting human potential.

[00:20:53] [SPEAKER_01]: Cody, I really appreciate you joining me today.

[00:20:55] [SPEAKER_01]: Thanks for having me, Dave. It's been a pleasure.

[00:20:58] [SPEAKER_01]: With as many breaches and security concerns as I report in this show,

[00:21:02] [SPEAKER_01]: it should be obvious that cybersecurity is not just about technology,

[00:21:06] [SPEAKER_01]: but also the human expertise needed to interpret and respond to complex threats.

[00:21:12] [SPEAKER_01]: Huntress is focused on elevating SMBs and MSPs around the world.

[00:21:17] [SPEAKER_01]: Huntress has a suite of fully managed cybersecurity solutions

[00:21:21] [SPEAKER_01]: powered by a 24x7 human-led SOC

[00:21:24] [SPEAKER_01]: dedicated to continuous monitoring, expert investigation, and rapid response.

[00:21:29] [SPEAKER_01]: And the proof is the execution.

[00:21:32] [SPEAKER_01]: Huntress is the number one rated EDR for SMBs on G2.

[00:21:38] [SPEAKER_01]: Want to know more about the platform?

[00:21:40] [SPEAKER_01]: Visit huntress.com slash MSP radio to learn more.

[00:21:59] [SPEAKER_01]: It's free and easy on the best way to support the show and help us grow.

[00:22:04] [SPEAKER_01]: You can also check out our Patreon,

[00:22:06] [SPEAKER_01]: where you can join the business of tech community at patreon.com slash MSP radio

[00:22:11] [SPEAKER_01]: or buy our Why Do We Care Merge at businessof.tech.

[00:22:16] [SPEAKER_01]: Finally, if you're interested in advertising on this show,

[00:22:19] [SPEAKER_01]: visit mspradio.com slash engage.

[00:22:23] [SPEAKER_01]: Once again, thanks for listening to me

[00:22:26] [SPEAKER_01]: and I will talk to you again on our next episode of the business of tech.