Avoiding the Cloud Money Trap

October 6, 2022

Season 1, Episode 2

When it comes to delivering in the Cloud, you need to know when to stay in your lane. People are the most expensive resources in any organization. Jon and Jay answer the question, what’s the golden path for secure application delivery in the cloud and how much of that can you outsource and insource while still remaining lean?

In This Episode, You Will Learn:

  • The significance of understanding your position when delivering in the Cloud.
  • The challenges and costs associated with human resources in cloud delivery.
  • The concept of the golden path for secure application delivery in the cloud.
  • The balance between outsourcing and insourcing to achieve efficiency.
  • The potential pitfalls businesses might face if they don’t set the right expectations and governance from the outset.

Themes Covered in the Podcast:

  1. Understanding Your Position in Cloud Delivery: Jon and Jay emphasize the importance of recognizing when to stay in your lane and the potential costs associated with not doing so.
  2. Human Resources and Cloud Delivery: The episode delves into the challenges of scaling human resources in line with business growth and the associated costs.
  3. The Golden Path for Secure Application Delivery: The duo discusses the ideal approach for secure application delivery in the cloud and the balance between outsourcing and insourcing.
  4. Setting the Right Expectations: The importance of setting the right governance and expectations from the outset to avoid potential pitfalls in cloud delivery.

Quick Takeaways:

  1. Cloud Delivery: The process of deploying and managing applications and services in a cloud environment.
  2. Golden Path: An ideal approach or method to achieve a specific outcome, in this context, secure application delivery in the cloud.
  3. Outsourcing vs. Insourcing: The balance between hiring external services or relying on internal resources for specific tasks.
  4. Economies of Scale: Achieving cost efficiencies by increasing production or scaling operations.
  5. Landing Zone: A set of best practices, design principles, and architectural recommendations for setting up an environment for cloud operations.
  6. SSL Certificates: Digital certificates that provide authentication for a website and enable an encrypted connection.
  7. Hybrid Cloud: A cloud computing environment that uses a mix of on-premises, private cloud, and third-party public cloud services.
  8. Infrastructure as Code (IaC): The process of managing and provisioning computer data centers through machine-readable definition files.
  9. CI/CD: Continuous Integration and Continuous Delivery/Deployment, practices that involve regularly integrating code changes and deploying them to production.
  10. Kubernetes: An open-source container orchestration platform for automating application deployment, scaling, and management.

Follow for more:
Jon Shanks: LinkedIn
Jay Keshur: LinkedIn
Jon & Jay’s startup: Appvia


0:00comes to be sort of measured upon right yeah I mean I have a bit of beef with this a little bit yeah just because

0:07security annoys me in a sense of security well lucky doors

0:20no one could stop me hello Welcome to Cloud unplugged this is

0:26season two episode four and I’m John Shanks and I’m Jake ashore and today we

0:34are mixing it up a little bit because we went a bit down the landing Zone um security uh rabbit hole for a little

0:41while um still in that guys um of obviously Cloud delivery assuming

0:48you have a landing Zone and those things um or assuming that you’ve got some level of scale I think we’re kind of still talking about companies like mid

0:54mid-sized businesses or Enterprises um

1:00a question for you Mr cashier where were you I’m joking

1:07um do you think it’s possible um given some of the decisions we’ve

1:13been making in a podcast because some of the com there’s a lot of complexity um around this stuff obviously I

1:19remember in previous episodes is it complex to somebody that understands it and you’ve got the answers maybe not as

1:25much but for a business there’s a lot of decisions that need to be made be it through somebody uh or a team or

1:31whatever because of that do you think it’s possible for a business to stay lean

1:37um and delivering the cloud successfully and support all the different business units accordingly yeah good question I

1:44think there’s probably a lot to unpack there and I’m gonna maybe slightly answer a

1:49different question which is completely different which is related which is how does a business stay lean um because

1:57obviously I mean the answer to can you stay lean yes of course of course you are uh of course you can all right so

2:03you can still even though it’s complicated even though it’s complicated you can stay lean because it’s it’s up

2:09to you how much responsibility to take on and how much you’re trying to solve problems that have already been solved

2:15right like you’re moving to the cloud so already you’re staying Lean by getting rid of

2:22um a massive part of your organization that dealt with physical infrastructure as an example so is it possible to do

2:31um application delivery in the cloud in a lean way yes now how does an

2:37Enterprise or a small to medium business try to try to do that well it’s it’s

2:44kind of about the services that they’re using in cloud and how they use that service those services and the knowledge

2:52that they have within the teams um so Landing Zone just as an example

2:57yeah we spoke about a lot of stuff in there and identity and then roles and you know

3:03which is like authorization stuff and then CI and like should see Ivy it’s

3:09like you know now doing infrastructure as code what’s the access for that thing and actually then maybe don’t use assess

3:14unless it’s got Runners that could be local or that the architecture of the CI systems are designed well and obviously

3:20better security posture and hop and spoke for networking and you know isolated Cloud accounts for the

3:27teams and environments and so there’s a lot of things a lot of buzzwords a lot of stuff right a lot of knowledge

3:34obviously that we’ve got over the years because we’ve been doing it for so long but um

3:39knowing all of those things it doesn’t sound like it is easy simple

3:45yeah it doesn’t sound like there is it’s very easy to stay lean based on those things so I get your you know what you

3:53um how you create a landing Zone like Landing zones are a known quantity nowadays right

4:00um you know you have a lot of uh companies that still will charge you hundreds of thousands if not millions of

4:07pounds to get you to just having a landing Zone and building it in the right way but you also have on the flip

4:15side of it you also have like um uh organizations that are building those things out all the time reusable

4:21products that can just give you it yeah instead of building something custom every single time yeah and

4:28um you know would are you staying lean in your costs by choosing one of them of

4:33course you are like the the the um a decision there is pretty much a no-brainer you’re getting the same

4:39outcome it’s a landing Zone um you’re making it really easy for it for a team that you’re on boarding in using best

4:47practice isolated accounts security all of that kind of stuff um now it’s the next bit of of right

4:54I’ve got a landing Zone how am I delivering my applications um what is that golden pass for secure

5:02application delivery and how much of that can I Outsource in Source how much

5:08knowledge do I need to have around that space to do it well um and some of that is I guess based on

5:17what decisions you’re making for the organization like are you are you allowing the development teams

5:25the full um you know the full freedom of choice to do absolutely everything that they

5:30want to do in which case you know you’re going to end up building something that’s potentially unmanageable because

5:37you’re trying to control all of the different ways of doing those things but if you’re being really strict on how

5:45um like you know um say making making it so that they’re only

5:50delivering containers and you’re going to orchestrate those containers in a way then you’ve confined

5:57um how they’re operating and you’ve put all of the controls in in the right place

6:02now what does that mean how do you how are you going to get like a you know kubernetes clusters or or something like

6:09that or something where they can um deploy and get environments

6:16um that’s the question yeah what is the thing that’s creating them environments where they can deploy something into I

6:22guess so just to pull on what you’ve said um and this is going to make us sound really biased I have to be kind of

6:28careful in the podcast but this is just me being sincere about

6:33it actually the most expensive things most of the

6:38time not always are usually people yeah and people are

6:44time and people plus time equals money yeah right so I guess either you’re

6:51Outsourcing and then the money could be higher because it’s a service and you’re Outsourcing that service and there’ll be

6:57lots to the management of the Service Plus how they’re managing the updates

7:02for you and then the skills that they have um so the price point might be way higher because of that obviously not

7:09necessarily a bad thing because you might not be able to have it in-house um but then there are things that are a bit more commoditized like you were

7:16saying there are products maybe or there’s Solutions off the peg when something’s quite

7:22specific right and general bit more general purpose like you were saying a landing Zone I mean we’ve kind of gone

7:28through this ourselves you can kind of make a lot of informed decisions doesn’t mean it will be 100 right for everyone

7:34there’s always going to be edge cases yeah um and lots of gotchas of like ah but you know X Y and Z so

7:42you know I suppose the real question is though how

7:48it sounds like the first thing to do is probably work out whether you’re us whether you’ve got a lot of custom

7:53requirements or whether you’re more general purpose and whether those General those custom

7:58requirements are going to manipulate the general purpose ones in a way that makes everybody succumb to the lack of

8:06commodity in it now yeah or whether you come deep up decompet mentalize how to

8:11speak those two things separately so you can say actually 85 of the business

8:17isn’t unique yeah just this bit is we’ll treat that separately and that becomes a

8:22separate cost it’s funny that you say that right um I I see that now a lot with people trying to do hybrid cloud

8:29um they end up you know um trying to make public and private

8:35cloud look like the same thing um but then yeah the lowest common denominator yeah exactly which is

8:41there’s no way you’re on Prem is efficient for that many services exactly

8:46right so you have 300 services do you have a man you have a number of

8:52people to support Community Services it’s probably not I have but then you might not need all those services to be fair I mean when you’re on you don’t

8:58need three when you’re on public when you’re on a database an object

9:03store done when you’re on um you know private Cloud you don’t really have Services you have people you have like

9:09you know Richard as a service or Peter as a service or John as a service you know it’s not but it’s the people I

9:15guess this is what I’m trying to work out just because I think it’s interesting to figure out because like you’re saying

9:22people are the Gateway for recognition of cloud yeah or products can be the Gateway for the recognition of cloud

9:29and if there’s a risk where unless you’re

9:34watching the spend closely you could easily sync a huge amount of

9:40money right which we see all the time on delivering to the cloud like obscene

9:46amounts like in the you know tens of millions really

9:51um if not sometimes hundreds of millions over the years isn’t it for large large Enterprises we have been there so yeah

9:59yeah so you know it doesn’t seem like the most cost effective way to go about something

10:07but it seems to be quite a common approach and I guess how to how does a business watch out for that so that you

10:13like what are the parameters to put in place to kind of help you not fall into that trap where Suddenly It’s like well

10:19it started off just being 500k and now it’s you know 50 mil I think I think

10:26um and this and this is uh you know it relates back to deliver or building anything right once

10:33if you’re building something and it’s not your domain it’s not your business domain like that’s not what you’re

10:40trying to specialize in as a business domain then you’re not going to do it efficiently and the cost to you to

10:47operate that thing that you’re building is going to be higher than um taking it from someone that does

10:54build those things um so let’s say uh even like a

11:00a a a really fancy meal right you’re not gonna

11:05um uh build a a whole kitchen just to have a fancy meal

11:11um you just want your fancy meal so you might go and um get it from a restaurant so there’s some something a service that

11:19a restaurant is providing you to give you your fancy meal you don’t have to build everything around that you know

11:25the operating cost for your kitchen the installation all of that just to get your meal no most People’s Kitchen no I

11:32suppose well I mean but we’re starting from basically okay so you don’t have a

11:37house you don’t have a kitchen you’re just trying to eat yeah I suppose okay I guess so at least a bit like I was about to say absolutely

12:21yeah I’m gonna cut this amazing meal it’s gonna be fantastic and you probably only think of what it takes to get that

12:29meal you don’t think of the energy the um the cleanup afterwards like all of

12:34the things that you have to do around yeah never give you the full expense I’m sold on the idea of eating exactly yeah yeah exactly it looked really good

12:42exactly exactly um maybe a better example still

12:48associated with food might be Pizza like not everyone has a pizza oven and you’re not gonna buy a pizza oven just because

12:54you want to eat pizza today right that day um so instead of having a kitchen it’s that’s it’s much better so but I suppose

13:01in this instance though the cloud they’re not just do it’s not a one-off so it’s like a continual thing it’s a

13:07continual investment they’ve made yeah and you know so they do there is an element

13:13where you could easily fall into the Trap by accident of

13:18it just the team that needs to support the business that’s making those decisions is now the team that needs to

13:26grow in line with the business to scale with the cloud so it’s almost like a

13:31one-to-one of like okay well if the demand increases that’s not scaled though right that’s to scale with the

13:37demand I’m saying not just yeah okay right because the demand’s gone up right internally yeah and now the team to meet

13:44that demand has to scale up yeah um and that’s then not efficient because obviously you’re scaling people in

13:50relation to exactly but like the definition so so that’s growing growing in line with the business you know that

13:57in in that sense is um it makes sense it’s when it’s when

14:02it’s human heavy you’re scaling up though the team presumably yeah scaling is used I think incorrectly most times I

14:10think in this case you’re scaling up the team to meet the demand like you would scale up anything to meet demand yeah

14:17but um you’re growing if you want to use that term because because scale

14:23um is is when you have the same number of resources doing more yeah right well

14:29that’s economies of scale um yeah I guess economies of scale is

14:34when yeah yeah you have proper economies of scale you get the same amount of resources because you didn’t need more people yeah but when you have just

14:42growth in a business growth in a business isn’t necessarily efficient

14:47because you you’re now growing everything almost at the same level which kind of implies that you’re not

14:53doing it to meet the same scale or getting true economies of scale in in

14:59that way does that make sense kind of I think that I think they’re different terms I think like getting economies of scale is one thing I think

15:06scaling you know like application scale right sure they scale up and they scale down you don’t say that’s not scale

15:12that’s growing yeah um it is that’s true scaling um and yeah you can scale people too are

15:21you getting the economies of scale no no I guess then then that’s the problem so I think

15:27the business could be scaling the demand is in like the increase of demand sure the team has to underpin that demand

15:34as your economies of scale within how that team’s set up so that it can basically not increase

15:41the cost of the footprint right yeah of that team to meet the demand like

15:46multiple times over I’m not sure whether people even think about that at

15:53the beginning because most of the time that team is quite detached to a point there’s a strong

15:58cost yeah from cost or from like you know it’s just it’s I don’t know it’s like it sounds like an expected thing

16:04somehow I think that’s the thing that people just seem to expect to spend a lot of money on those things maybe I’m

16:11not sure um but it does seem weird that it’s not like you’d want it to be lean and as

16:17cost efficient as possible it’s funny because it sort of ties back to the very first episode when we were talking about

16:22how you’re measuring teams and devops and you know all of that kind of stuff cost wasn’t even a factor cost is not a

16:29fact it’s never a factor is it no it’s like how long does it take to recover what’s your change right like how fast

16:36do I mean how fast did you get to production how that’s all you measured does it cost you yeah then you’re doing it well yeah yeah it could cost you

16:42eight bazillion pounds yeah it’s a long time crazy yeah it is a little bit crazy in

16:50that sense because it’s a I guess that’s at the engineering level of what you’ve implemented how

16:56much it costs well if you’re an engineer then you’re probably not thinking of it because it’s a song cost you like you

17:02are there you’ve got a job you know the business yeah right exactly that’s not your responsibility is it I guess if

17:08you’re an engineering manager then you’ve got a bit more but you might not have control of that budget but if

17:15you’re a head of Department you certainly do yeah um but then they’re a bit detached from

17:20they might be a bit detached from the outcome um that’s being like and weighing that

17:26up so my department cost me this much this is how much they’ve achieved in this time and how that is growing yeah

17:32well it’s actually different because we spoke about this before there’s usually three buckets there’s the cost of

17:38delivering it right now yeah like implementation cost there’s a maintenance cost which might

17:44still require a lot of domain knowledge depending on how complex the thing is right so then you can’t really have that

17:50team has to remain because they understand it the most yeah and then you have a support cost which is like and

17:57that could be the same team sometimes sometimes it could be a separate team but it’s a budget itself delivery

18:03maintenance isn’t maintenance it can be bundled together sometimes I’m not saying it’s always but I mean the cost

18:10of maintaining that thing might not be the same cost as the number of tickets coming in to support it yeah upgrades

18:15patching that type of stuff maybe it’s Auto patchy maybe it’s autocoding I don’t have a maintenance cost if it’s all I’m I’m I kind of think it is

18:23slightly separate so like there’s an initial cost which could be like a build cost yeah yeah and then you have a a

18:31cost to just support what you’ve built

18:36um and that is like the you know the upgrades patching all of that kind of stuff of the software but then

18:42presumably your thing has users so now you have a kind of front-facing user

18:49cost um to how you know more getting more demand because then you’ve got like

18:57any one of those things could be more uh it could be cheap or expensive yeah like

19:03um the build you might put a lot of that up front um cost into build so that you are what

19:09you’re telling that one cost yeah exactly so what you’ve ended up building yeah could also be very expensive to

19:15maintain exactly exactly it could also be very complex to support to onboard your new users on and all that which may

19:21which may be worth it I’d write based on where like where you were and what you managed

19:29to achieve yeah but I think you know nowadays I’d say that there’s probably more

19:35efficiency possible you know like the industry’s moved on definitely um without a shadow of a doubt yeah

19:41exactly there are just more efficient cheaper ways to do things like like we

19:47were saying earlier with um uh you know um SSL certificates

19:53yeah exactly just like that do I need a web dive to hosting so many kind of commodity Services both

20:03in terms of product and like service that you get from places where it’s just like okay yeah it is cheaper for you to

20:10make me a pizza and I could buy it for five pounds rather than me getting it yeah getting all the stuff and the time

20:16and energy exactly exactly yeah because just don’t do it don’t make your own

20:21pieces pizzas people like go to Pizza Express yeah it’s funny that it’s a funny one

20:27because it’s a tricky because the things we were saying it’s a hard

20:32I think the right thing the answer has to be you need to put the right like expectations and governance in place at

20:39the beginning so that you’re measuring stuff fairly off on the offset I don’t

20:45from my experience it’s it’s that’s quite an advanced way to think yeah exactly and you’d have to have gone on

20:51the journey probably already and really understood it to think in that way um to realize like the pitfalls that

20:57could happen I guess you know if you’re because because nowadays

21:02um everything is usage based right like the the common pricing strategy for for most things is usage based right cloud

21:09is usage based I’m sure there might be an initial cost for you to like have start having those conversations and and

21:15do things commercially but one thing that isn’t really massively

21:20usage based is people like once you’ve hired those people you have those people right you’ve now got a duty to those

21:27people and everything else so like you could you know be a bit bold and try

21:34something buy something that is specializes in this space and and and

21:40maybe it will meet your requirements or you can feed into that thing to make it

21:45meet your requirements rather than buying people to build from scratch to do something

21:51with that might maybe meet your requirements but then you have a massive cost to maintain forever

21:57um yeah because it’s a trick it’s a tricky thing because the Cloud’s the platform in itself yeah however it’s got

22:03no real opinions right and there’s loads of services and to support lots of different businesses

22:09that can’t afford to have loads of different opinions so then you need to have opinions on on

22:14that on those things inside of your business and some of the businesses if not most

22:20can roughly have roughly the same requirements like we kind of went through everything today so in theory

22:27the cloud vendors because of the level of scale they’re up might not be able to make those decisions for you because we’ve got such a variety of sizes of

22:33businesses however there are sensible decisions that you can just make they’ll just make your life easier like which is

22:39kind of did on the previous podcasts um but then you also need someone to help

22:45recognize all of that value which means then you need to hire somebody um and then you’re not going to be sure

22:51like is there a vision or a strategy for this right so it’s like because at the moment the vision’s cloud it probably

22:57just stops though right so like you deliver on that Vision once and it’s not like yeah but that’s it they might not

23:02need of it might not feel that they need a vision for the delivery off Cloud yeah yeah right so it’s like well I didn’t

23:07know it like I’ve gone to the cloud we’ve got apps apps Cloud just smile yeah people will just come in and make

23:14this work join these two things together yeah so that I can get it and get what I want which is like certainly higher and

23:21then it’s like well you know a bit like a builder this is quite complicated I’m not sure like looking around making notes this one will need to come down

23:28and you ask for this and then you’re like oh my God this is really like way more than I thought it was going to be

23:35um so it is difficult I think but there definitely are I think you’re right there definitely are much old more

23:42alternative routes than always having to scale people in relation to it nowadays and I think the building buying thing

23:49has been a paradigm throughout forever for like literally forever yeah people used to you know

23:55um Even build like PCI and all of those things right that that used to be

24:00and still is to a certain degree massive but now you can just Outsource your

24:06payment information to a service and then you don’t have to deal with that headache like now it’s just paid for

24:11PayPal that’s exactly what exists yeah yeah exactly so it’s just like um

24:18like it feels like we’ve you know in in in our thinking evolved a

24:24little bit to know um that there are Services out there um that deliver a really fixed

24:33um solution to something that you can just go and use and it’s more efficient to use it in that way because it’s usage

24:40based then having to build it yourself once you’ve built something you’ve got to make you

24:45just got to maintain it and then the risk of not maintaining it properly

24:50um and not you know meeting kind of user needs constantly is just so high it’s

24:58just like it’s so high I don’t keep on top of the industry the way that it moves exactly it’s like that team needs

25:05to be r d in yeah right yeah working out ahead of all the other teams like how

25:10much more efficient ways of delivering with the way you’ve implemented plus now managing maintaining and supporting user

25:17research what you’ve just done product management yeah because r d essentially it’s an internal product in some degree

25:23yeah definitely or internal Services wrapped with lots of weird custom things trying to be a product that isn’t

25:30necessarily seen to be a product um so yeah so you end up in like the

25:35kind of common platform teams but then they’re not necessarily proper platforms all the time they’re like kind of just

25:42strung together yeah maybe ways like basically stringing together like bits of automation around other things yeah

25:49it’s who they’re built by yeah you know I guess I’ve got a little bit of a bias

25:54opinion yeah because it’s kind of we it’s kind of where we’ve come from right like

26:00um and and I’m I am you know that uh you know role as well have been in the past

26:05where um devopsy type people they they come in and we are used to stringing things

26:13together well solving problems solving yeah totally it’s coming to solve a problem 100 we’re saying you see the

26:18problem yeah you use your knowledge you look at the industry and you solve the problem exactly and that’s what you do

26:23yeah you’re solving the problem um but the way that you’re solving the problem is normally but by putting

26:30things together I.E other people have solved problems and you’re gonna solve the overall problem for the for the

26:36business by putting all of those things together now there are and and that

26:41might some you know evolve evolve evolve um but those things that you have to put

26:48together there’s now solutions that make it really efficient um where they’ve solved it like git take

26:55GitHub or gitlab as an example um Source um version management

27:01um used to be completely separate from CI and CD Now together simplified a lot

27:07of things you don’t have to combine those things and that’s now buy yet another product buy another product for

27:13someone to to go and string together to to solve a business problem is just buy a thing because it sounds little um so

27:21just like that there are like operating in Cloud there are definitely things that you can buy that have those things

27:29bundled together that you can just get the outcome really quickly yeah

27:34um but also when because it’s just rolling back a bit the people that come in to

27:40solve those problems they’re looking at each problem as it surfaces

27:46a lot of the time yeah right and to have stepped back and then thought through

27:51where you’re probably going to end up ahead of time and then come up with a vision to

27:57protect the business so it is going to be more cost effective and it is going to be secure and it will be able to give

28:04me economies of scale with the team or whatever is important from a business objective and it could be like plethora of different things but

28:10very very rarely happens in a team like that because they’re being more reactive

28:15to the decisions made somewhere else and then are dealing with the problems as they’re servicing and then making

28:21tactically strategic decisions to a degree like and then kind of just iterating along with the iterations that

28:28they’re surrounded by using industry tools that they’re surrounded by yeah

28:34and so there isn’t necessarily A Vision it’s just a way of working um yeah exactly and that can be really

28:40yeah it could be really inefficient yeah it could be really expensive it could be really insecure like you know there are

28:46principles that try and prevent it obviously like we mentioned but um unless you’ve thought about what you

28:51want from the teams that are going to enable you then yeah you’ll just end up that’s just exactly what will happen because it’s

28:59designed that way and it’s no one’s fault because that’s the yeah it’s just how

29:04it’s how you begin on the journey that’s what you start with then you get someone in and then they’ll be like actually can

29:09you also help with the hiring for someone else and then they go and get someone they used to work with maybe or whatever kind of goes on and then so you

29:16know fast forward a few years you then like do people even like because you know like you were saying it’s kind of

29:23you have you build a team and sometimes you build a team by accident to solve a problem that you didn’t think that you

29:28would be trying to solve yeah and to solve those problems

29:35well you know like you kind of um touched on there you have to have a lot of

29:40different skill sets yeah and constantly be doing that like user

29:46research product management industry knowledge yeah industry where the Market’s going everything everything all

29:53the time and yeah that costs that it’s a huge huge and and the time the time and

30:00the ability inside like to retain that person that can do that is probably highly unlikely God in this environment

30:06yeah it could be worth the fortune yeah if they could navigate a business to some really good place yeah so yeah it’s

30:13it I I don’t I think these products and tools out there that’s just been sensible about

30:19which ones Drive the most value or Roi basically it’s like yeah what’s going to give you the return on investment here

30:25do I not need to scale up the team as much do I kind of move just as quick because it’s already done like there’s a

30:30load of work it’s never there’s no one product that solves all problems obviously so you have to look at each

30:36thing you’re trying to do slightly independently and they’re doing a bit of Investigation tonight is there already a leg up in

30:43this um is what I’m doing gonna actually meet demand and scale if I anticipate there

30:48to be is a way of working conducive for the scale that I might hit and the demand that might hit

30:54you know in those things you kind of have to look at um before you go too far because then you

31:00have to revise that’s kind of um like what the you know Cloud Center of

31:05Excellence Enterprise architecture type role is does it fit the business and its

31:10needs and its uh potential scaling um with that requirement you know with that

31:15with that thing um makes sense um but then in in terms of like

31:23um how it sort of um evolves over time how it’s measured

31:29um the we touched on it briefly in the last episode but like security of the

31:36thing that you’ve built um that’s never something that necessarily gets baselined and then

31:43constantly sort of measured upon right yeah I mean I have a bit of beef with this a little bit but yeah just because

31:52it annoys me I’ll just say it no it’s security annoys me

31:58in a sensor security well lock your doors security sorry let me just rephrase security

32:05doesn’t annoy me just generally lock all your dolls I’m coming for you no one can

32:10stop me um I think it being a post event thing

32:17frustrates me right like hey I’m going to attach a thing that’s going to tell me everything that everyone’s done wrong

32:23and to then not going behind well clearly whatever it is you just delivered didn’t work right so he’s like

32:30why are we pretending that it was all really good but it clearly wasn’t because it was really good unless it’s a security thing and everyone said you’re

32:36just attaching it as a bit of an insurance policy great fine makes sense right check the things when they happen

32:43just that insurance policy I need it I need to know those things I get it when it’s used in that sense

32:48but given some of the like Flex air reports and security being the biggest challenge like the highest Challenge and

32:54then all like the market and these security products is definitely telling you that cloud delivery is not going

32:59well right it just isn’t going well right which then begs the question of like who’s doing it how are we

33:05delivering it who’s looking at because clearly it’s slightly flawed there’s no point pretending like if the data wasn’t

33:11there fine and if everyone was like putting in external products and they’re like yeah good job

33:32there is a joke

33:43I have no idea whether we’re in a secure position or not we need to get something right the risk is too high nobody can

33:51tell me the answers of how we’ve actually done Security in the delivery I’ve got no way of looking to see like

33:56this you know there’s like different every time I speak to somebody and ask questions it’s really hard work to get

34:01the answers and I’ve got lots of projects that have delivered and here realize the thing around bloody been scaling properly in a consistent way and

34:08making sure these requirements come in because you can’t wait until you’ve delivered and spent loads of money and

34:13then be like oh well we have to spend some more money yeah exactly yeah we’ve got to check it all now but yeah and

34:20then we’ve got to try and somehow fix it and even know how we’re working is wrong isn’t working yeah

34:26so it’s like so what are we fixing it with oh you know just policy as code what that’s not a fix yeah exactly so I

34:32mean that’s not a proper fix it’s a part of a fix you know and it’s it’s a it’s a principle it’s a standard it’s not a

34:38solution so it does get me a bit annoyed when people just don’t call it out and

34:43you’re just like clearly can we just agree it’s a problem I’m not saying everyone’s doing a bad job then it’s

34:49really bad and you know but there’s clearly enough of a problem there for the reports and the data to tell us that

34:54there’s a problem I mean it’s just you can’t so can we just we just need to

35:01have the right conversation to fix it because it’s clearly not working right and that’s what I’m trying to do I suppose if the questioning is

35:07it is a very tough problem to solve and and therein lies the issue it’s not like

35:13magical there’s your answer but there are solutions and things trying to address it it’s like you don’t have to

35:19solve it all right and there’s multiple ways of solving it right like just before we started we were talking about how people

35:26are kind of um uh get into cloud and delivering

35:31applications and you know people if you’re if you’re just like creating a website

35:36um for your company you don’t have to build a website from scratch from code you can use something

35:42like WordPress or Wix or something like that yeah right so you’re all ready space all those yeah exactly so you’re

35:47already getting to cloud in some way in a way that makes sense for the size and

35:53scale of your business um but kind of by accident because you’re just using a service you know that that service is on cloud but you

35:59know you’re you’re doing it so um uh just just like you know if you’re

36:05trying to move to to put public Cloud really what you just want to do is

36:11figure out a way to deliver fast like cost effective securely and you know

36:19with the right flexibility to allow your organization to move with its

36:25requirements over time yeah right then you’re going to go to public Cloud um but who’s going to build it for you

36:31how are you going to get there you know what you’re going to buy exactly that’s the problem isn’t it yeah that’s the

36:36problem because it’s like 50 different Million Ways to do it 100 um some of them are quite common but again

36:42those things I think are fine but the tools the tools and people it’s a tool and people industry in that space which

36:48means it doesn’t have economies of scale for that reason yeah it’s not designed for it it was designed to solve a

36:54problem for that Persona it wasn’t designed to solve a business problem yeah right those tools fine it’s fine

37:00there’s nothing wrong with that yeah those tools were designed to make that person’s life easier that was delivering for the cloud yeah

37:06um totally different proposition but then it’s like but actually if I do one economies of scale with the people

37:12then what are my answers and I think that’s another lens that you kind of have to question and there are things out there and I know we’re not gonna I

37:17don’t want to bias the podcast too much or like reference lots of other companies with people but I think the answer is

37:24you know you should make sure you’re really hitting on what’s important to you in the business early yeah

37:31um and then go and do the research you’ve got to find your way to the solution you’ve got to find your way to the solution yeah exactly and then

37:38you’ve got to also make sure that you set caps I think some industry benchmarks like yeah what’s the ratio we

37:44should be aiming for for the delivery teams to this Central team if it is Central or if it’s not Central what’s

37:50the ratio of the number of people that could support the number of projects at once yeah um you know what’s sitting behind those

37:56people what exists in the industry to set that up how do I get the right behaviors of that team it’s Downstream

38:03do it does this person have to keep communicating backwards and forwards between everybody what happens when

38:09we’re three years down the line you know am I gonna have a big maintenance cost What’s the total cost of ownership of

38:15all that’s been delivered and I think they’re all the right questions to ask early exactly because then you can make

38:21informed decisions on like what you want to do and it might be that spend loads of money on people people and what if you’ve like scaled up your team to meet

38:27some demand but then your business changes and you don’t have to meet the same demanding again you still have to meet that Demand right

38:34so do you are you allowed to like just remove people from your team yeah

38:40and then you’ve got all that headache exactly that’s like yeah the change is huge flip side isn’t it and it might be

38:46that you’re trapped because you’re like well exactly with this system’s supporting like we might not have grown

38:51anymore yeah but it’s supporting like I’ve now hired 10 people yeah and they’re capable of supporting I don’t

38:57know you know a thousand apps or whatever exactly these people anyway yeah because of the

39:05cost yeah of what’s being delivered so it’s not commodity and it’s not mainstream it’s not off the shelf exactly I don’t go to another company

39:11and just get like some additional I don’t know vcpu tax yeah right it is

39:16like a tax on something because that’s how they charge but I get support they own the technical risk exactly that

39:22um you know in delivery of it instead I own it all now as a business every little thing that we’ve done

39:28I know maintain a known it’s just it is I think it was fine a fair few years ago

39:34when it wasn’t as complicated but I just feel like it’s got this what you’ve got to do now in the landscape

39:41to meet those needs of the things that we spoke about for the business means you have to really think about commoditization now whereas before I

39:48don’t feel like you had to as much because hey people didn’t have loads of accounts but

39:54I mean years ago it’s like that wasn’t even really a thing people would say in which case you kind of were sharing yeah

39:59but also it wasn’t even really audit that much when we started it wasn’t even a service right properly and when it was

40:06It was kind of in the account then later on it was out the account and so it’s like yeah things have evolved massively to

40:12support the scale of the market overall um and the opinions of the market

40:18overall yeah because that’s the thing that has enabled like that’s the reason that those Services now exist the

40:24opinions were that hey it’s not good to have shared accounts we need to go and

40:29you know you need to have isolated accounts right great great opinion great um uh standard to have how do we do

40:36that thing yeah how do you make those accounts which people did do yeah but it was really like automated manual I think

40:44from memory related manual yeah there’s a level of automation that can happen after you’ve done it all manually first

40:50all right script that does something well yeah versus buying a product yeah yeah

40:56already automated but there’s nothing manual to do yes right so but when it’s internal it’s like right well we’re

41:03going to automate eventually yeah but first of all I need to write all this stuff yeah before we then can automate

41:08around all that stuff

41:14yeah exactly so you know it’s kind of manually automated that’s what I call it

41:19it’s a bit of a bad term but you’re not obviously moronic northern states

41:25contradict myself the entire time say one thing that means something completely different uh I’m an automatic

41:31manual yeah exactly so exactly Steve’s got very manually automated to me

41:41so anyway I guess we haven’t really massively solved this we I think I don’t

41:47think I mean we tried our best but is it complicated it’s a complicated thing the you know tldr that the things to look

41:54out for when um kind of starting to own a domain is

41:59how much it’s really going to cost you how effective you’re going to be and how quickly you can be effective yeah and

42:06and trying to do that with a a lens of the initial the initial thing to solve

42:12the the business problem and then over um you know the time that that solution

42:18is going to be alive for yeah and if you do Outsource those things because they do have the experience and you don’t

42:24yeah then tell them to tell it tell you well how long does it normally take you

42:29to do this repeatable thing that you always do because it should be short if you’re doing it all the time and if

42:35you’re in this like well it depends it’s going to depend on your business at the age of that you just think okay you’re being sold exactly though you know does

42:42it depend I mean obviously there are like I think the art is to say just give

42:48me your general purpose thing first yeah exactly and then we’ll have a then we’ll have a conversation about what’s custom

42:53to us yeah but at least then you can like time box the commodity aspect of it

42:58to not fall into the Trap of yeah understand what the success matter is yeah exactly then like break it down and

43:04don’t fall into like the oh well it’s going to cost is like all this because we’ve we need unique things yeah exactly

43:09um it’s like they’re the things that shouldn’t like they will cost but you can deal with those things separately but anyway 100

43:15cool right well bit of a ramble and obviously had a bit of a rant yeah and loads of oxymoronic terms I think that’s

43:22the most uh passionate we’ve seen the automatic manual jobs

43:28um cool, well stay tuned and we’ll be back in our next episode I’m not sure what we’re going to talk about but we’ll

43:34come up with a topic if you have any topics you want to hear us talk about or

43:39you disagree obviously again feel free to reach out great to speak to you again bye bye


43:56thank you