Ray Mota, Host of The Deep Edge Podcast

ACG HotSeat: Juniper Networks Michael Bushong discuss the SP/Cloud/Hyperscaler space

Industry Voices Cloud MetroOperations
Ray Mota Headshot
The still image from the podcast video is of Mike Bushong, Group VP, Cloud-Ready Data Center, Juniper Networks, on a Zoom screen during the interview talking to Ray Mota, Host of The Deep Edge Podcast.

Juniper’s Bushong gets real about the complexities of the service provider/cloud/hyperscaler relationship.

In this episode of the Deep Edge Podcast, host Ray Mota talks with Juniper’s Michael Bushong about the cloud/Hyperscaler relationship (are they partners or competitors?) and discusses how service providers (SPs) should best think about the distribution of profit pools. Hear Michael’s take on horizontal vs. vertical offerings, as well as the role Juniper plays — and will play –– in the hyperscaler/SP relationship.

Show more

You’ll learn

  • How SPs can maintain control of the billing relationship with the customer without damaging the hyperscaler relationship 

  • Why Juniper prefers and provides horizontal platforms

  • How Juniper gives SPs the confidence to move forward successfully in the cloud/hyperscaler space 

Who is this for?

Network Professionals Business Leaders


Ray Mota Headshot
Ray Mota
Host of the Deep Edge Podcast

Guest speakers

Mike Bushong Headshot
Michael Bushong
Group VP, Cloud-Ready Data Center, Juniper Networks


0:00 you're now listening to the deep edge

0:03 podcast with your host technology expert

0:06 rain Moda giving you the rundown on

0:08 Telecom cloud and all things business

0:11 and Tech this is the deep edge podcast

0:16 hello hello and welcome to another

0:18 episode of the deep edge podcast and

0:20 please note that in this particular

0:22 episode we'll be Distributing this in

0:24 video and audio and your favorite

0:27 platform and I'm excited that joining us

0:29 today we're very honored to have Mike

0:31 Bouchon who's group vice president for

0:34 cloud ready data centers at Juniper

0:35 Networks Mike thanks for joining the

0:37 podcast

0:38 absolutely my pleasure it's good to

0:41 spend time together again yeah exciting

0:42 that you were able to get me during

0:44 baseball season just when the playoffs

0:46 are about to start so good timing

0:48 it is my team is in the playoffs that's

0:51 right both our teams are in the playoffs

0:53 so I guess we're all in a good mood now

0:54 Mike maybe before we get started could

0:57 you tell us a little bit about your

0:59 journey and how you got to your current

1:00 role at Juniper

1:02 uh sure I mean the the short story is

1:05 started uh designing uh jet engines in

1:09 college became a tech writer I mentioned

1:12 that because it's kind of an awkward

1:13 origin story went through the years at

1:16 like Silicon design companies moved into

1:18 networking found out I'm pretty good at

1:20 strategy uh bounced around tried to make

1:23 my fame and riches in sdn and then

1:25 ultimately ended up back here at Juniper

1:27 where I lead our data center business

1:28 all up which covers off you know

1:30 certainly hardware and software and then

1:32 obviously the space we'll talk about

1:33 today

1:34 this is excellent I I know you for all

1:37 these years and I didn't know that part

1:39 of your background so that's uh thanks

1:41 for sharing that well maybe we'll just

1:43 start off with what what are you hearing

1:45 from the SPs

1:47 so we focus a lot on the Telco Cloud

1:49 space

1:50 um they're looking obviously at how they

1:52 roll out new services and you know it's

1:54 hard to to talk about anything in the

1:56 cloud space without mentioning the

1:58 hyperscalers

1:59 um I mean there's real questions right

2:00 our hyperscalers are they partners are

2:04 they competition and so what everyone's

2:06 trying to figure out is you know what's

2:08 the extent to which you cooperate or

2:10 that you compete with these types of

2:12 companies

2:13 um you know how do you structure your

2:15 strategy so that you can control your

2:16 business what are the routes to Market

2:18 going to be what are the products and

2:20 and you know underlying service

2:22 implications and then ultimately how

2:23 does that kind of break out into the to

2:25 the business where you start looking at

2:26 the division of profit pools I mean that

2:28 was really kind of the Crux of the

2:30 conversation

2:31 yeah no I mean I think that's that's

2:33 been a Hot Topic uh for a while then I

2:36 think some of that is from a

2:38 hyperscaler's point of view uh doing a

2:40 better job of having to show the

2:41 visibility and their plan on how to

2:43 better partner and in some cases it

2:45 might be a coopetition type strategy who

2:48 knows right so it's it's interesting

2:50 here but what what did they specifically

2:52 say the direction about you know their

2:56 relationship with these hyperscalers

2:57 moving forward do they do they feel

2:59 there's uh an important role in what was

3:02 the driver for that

3:04 if you if you essentially concede your

3:07 business to the hyperscalers and put

3:09 everything in the hyperscale space you

3:11 lose control right you lose

3:13 um ownership of the accounts

3:14 um there's actually service and

3:16 performance implications if you're just

3:18 using you know hyperscale points of

3:20 presence

3:21 um if you look at where the data centers

3:23 are things like performance still matter

3:24 and so you know how do you meet you know

3:26 individual customer requirements if

3:29 you're beholden to somebody else to

3:31 create the the data centers that surface

3:33 those those applications

3:35 um the big question though is like who

3:37 whose business thrives in that model and

3:39 who maintains control some some people

3:41 have like you know very very firm ideas

3:44 of what strategically they're going to

3:45 go and other folks are probably earlier

3:47 in their journey in their Discovery

3:48 process and it's interesting to see like

3:51 just how how different people are and

3:54 and how they're looking at the problem

3:55 space and it leads to kind of wildly

3:58 different perspectives on I mean we

4:00 really are like kind of in this this

4:02 phase where it's you know what are the

4:03 right decisions to make you know how do

4:05 we navigate you know the current

4:07 economic climate you know especially in

4:09 Europe by the way where things like the

4:10 US dollar trading at you know all-time

4:12 highs against local currencies you know

4:15 what does that mean for for capex

4:18 yeah no I think that's uh that's a Hot

4:20 Topic from that perspective and one of

4:23 the things you mentioned is about

4:24 different angles uh from that regarding

4:27 type of latency requirements I mean when

4:30 you look at the whole thing about you

4:32 know private Cloud right is there are

4:35 some potential advantages there and how

4:37 are SPS navigating to plan that we hear

4:40 this whole discussion about creating

4:42 some type of vertical or horizontal

4:44 offering yeah so the I guess generally

4:47 if they look at the overall Services

4:48 there's two models right you could have

4:49 a vertical vertically oriented or

4:51 vertically integrated stack or you could

4:53 have a horizontal stack where you look

4:55 at like a traditional platform play and

4:57 I think there's a little bit nervousness

4:58 in the big vertical Stacks primarily

5:00 because you lose

5:01 um economic leverage if you are bought

5:04 into like a single stack then they own

5:06 you know that that vendor that supplier

5:08 owns end to end and if that's the case

5:10 then the challenge is that the pricing

5:12 that you enjoy in year one is not the

5:13 same as the pricing you enjoy in year

5:15 five and if the transitions are

5:17 operational it's not merely deploying

5:19 new equipment I mean you're talking

5:20 about retooling you know entire

5:22 processes you know re-qualifying

5:23 Services which is typically a multi-year

5:26 Endeavor and so if you lose pricing

5:27 leverage there's actually no easy exit

5:29 and so there's a lot of discussion I

5:31 would say actually rough consensus that

5:33 that if they can create

5:36 um you know freedom from a you know

5:37 using a horizontal platform and then you

5:39 can use either best of read or at least

5:42 create a little bit of competition in

5:43 the individual service components that

5:45 that gives you kind of economic control

5:47 over your fate which you know obviously

5:49 if you're staring at uh you know margin

5:51 pressure generally as a at a business

5:53 level if you're looking at macroeconomic

5:55 headwinds maintaining some control over

5:57 your fate seem to be

5:59 um again a pretty widely popular opinion

6:03 on in terms of where to start

6:05 yeah I mean I I think this this is great

6:07 because I think you know there's a lot

6:10 of discussions about certain

6:11 architecture but not enough discussion

6:13 about the impact on the business right

6:16 um uh side of it what what do you feel

6:18 that SP should do about their whole

6:21 distribution of the The Profit polls

6:23 there

6:25 um I think you have to understand or at

6:28 least develop an opinion about how you

6:29 think the profit pools will be divided

6:31 between you and the partners I think

6:34 there's probably three dominant models

6:35 uh one and you can presume that the

6:38 service providers maintain control of

6:40 the account and that the essentially the

6:42 the hyperscale part the cloud part of it

6:44 is is merely part of this the service

6:46 that obviously gives the SPs control

6:48 over the account it gives them control

6:50 over the service

6:51 um it gives them the ability to decide

6:54 how the profit pools are divided up the

6:56 other model is that you know it's like a

6:57 meet in the market what they're going to

6:59 do is you know Enterprises will

7:01 essentially use cloud they'll use

7:03 connectivity services from the SP side

7:06 and in that model the idea generally is

7:09 you know people purchase what they

7:10 purchase and the

7:13 um and the pricing will kind of dictate

7:15 how the profit pools are divided I think

7:17 that's actually a pretty dangerous

7:18 position to take right if the thing that

7:20 they're looking for the Enterprises are

7:22 looking for is to move a workload to

7:24 Cloud then the the dynamic that'll

7:27 dominate pricing and really the

7:28 distribution of the profit pools will be

7:30 how the cloud is priced so if there's a

7:32 hundred fictional dollars to spend and

7:35 the cloud in that in that situation

7:36 let's say the hyperscale charge is 80

7:38 for their service the SPs can't go back

7:41 and say well my service is worth 50

7:42 they're going to get whatever's left

7:43 over so even in the meet in the market

7:45 model where it's the kind of friendly

7:47 coopetition if you will

7:49 um they actually lose control in that in

7:51 that particular model then of course

7:53 there's the the cloud model which says

7:55 that you know the connectivity service

7:57 is merely a part of the cloud service in

7:59 which case the the pricing is fully

8:01 dictated by you know the cloud the

8:03 hyperscalers who own uh the relationship

8:05 with the customer so obviously in that

8:07 model that's a that's even more dire the

8:09 point of all of this though is that even

8:11 that middle one which is kind of where

8:12 people they tend to think of that as

8:14 safe neutral ground I don't think that's

8:16 actually going to be safe in neutral

8:18 ground which means that the service

8:19 providers need to be thinking today how

8:22 do they construct their services so they

8:24 can send actually you know maintain the

8:26 the billing relationship with the

8:29 customer because that gives them control

8:30 over how that the dollars are going to

8:33 be distributed and if that's the case

8:35 then what's the value that they have to

8:37 provide in order to make that that

8:38 relationship stick and it's not just

8:40 providing connectivity I think you have

8:42 to think through a little bit like

8:43 what's the role of the service provider

8:45 there

8:46 um and then there's certain competitive

8:48 advantages that the service providers

8:49 need to lean into certainly if you look

8:51 at the the hyperscalers they've got um

8:54 you know data centers but it turns out

8:56 like as you mentioned are things like

8:57 performance matter things like latency

8:59 matter

9:00 things like the location of your data

9:02 center actually matter and so if you've

9:04 got all of this real estate how do you

9:06 leverage that to go and dictate the

9:08 terms under which you're going to

9:09 partner with the hyperscalers I think

9:11 there's elements of that that can be you

9:13 know really transformative in how you

9:14 forge those Partnerships yeah no I think

9:17 this is great you did a good job

9:18 breaking it down into the different

9:20 components maybe we can close with a

9:23 little bit about you know from a juniper

9:25 point of view I mean what is uh Juniper

9:28 strategy and position regarding this

9:31 uh so you know obviously we call the the

9:34 largest service providers customers you

9:36 know they are all using and have been

9:38 using Juniper stuff for a long time

9:41 um we provide horizontal platforms we

9:43 believe in choice and flexibility

9:44 fundamentally we believe that unlocks

9:46 some economics and performance things

9:48 that need to that need to play into this

9:51 so I think we generally want to help you

9:53 know the service providers maintain

9:54 control we also obviously have you know

9:57 great Partnerships with the hyperscalers

9:58 we do a lot of business with all of the

10:00 cloud companies and so you know the the

10:02 question is what are the technologies

10:04 that have to bridge between these and we

10:05 think that we can provide those

10:06 Technologies the thing that I would say

10:08 strategically though right what we want

10:10 to do is

10:12 um help folks navigate the overall

10:14 transition yeah I don't think this is a

10:16 pure technology transition I think

10:18 people that look at this as you know an

10:20 Opex or capex exercise that look at this

10:22 as a protocol or a technology and they

10:25 reduce it to just execution I think

10:27 they're they're missing the bigger game

10:28 here and I think the role that Juniper

10:30 ultimately will play is that we're in

10:32 this unique position to see kind of all

10:34 sides of the problem space which allows

10:36 us to act a little bit like a consultant

10:37 in the space and talk a bit about what

10:39 are the things you should consider I

10:41 won't say that we always have the right

10:42 answers for everybody indeed you want

10:44 companies to to own their strategy

10:46 themselves because they have to arrive

10:48 at those decisions but we are in a

10:49 unique position to help surface the

10:51 right questions and then those questions

10:53 and certainly the answers to those

10:54 questions have implications on the

10:56 underlying technology I think if you can

10:58 Plumb that through then you can get

10:59 clear line of sight which should give

11:01 service providers what's missing is the

11:04 confidence to move forward deliberately

11:05 in a particular direction if we can help

11:08 convey that confidence then I think we

11:10 put people in a better position to

11:11 execute yeah no this is excellent I'm

11:13 going to give you a round of applause

11:16 for my audience there but Mike uh thanks

11:19 for joining the podcast hopefully you'll

11:21 join us again and give us an update in a

11:22 few months

11:24 of course well with Mike this is Ray

11:27 motor thanks for joining this edition of

11:28 the deep edge podcast

11:32 thanks for joining us for this episode

11:34 of The Deep bench podcast make sure to

11:37 follow Raymond on Twitter and Linkedin

11:38 that's our Moda if you enjoyed the

11:41 podcast please make sure to subscribe

11:42 rate and leave a review on iTunes want

11:45 even more head over to

11:47 acgcc.com to access exclusive bonus

11:51 content till next time

Show more