amuck-landowner

Why people pay so much for big brand VPS and cloud

graeme

Active Member
There have been a few threads that discussed 1) why people pay a lot more to use AWS etc. and 2) the likely consolidation of the VPS industry.


Looking at this from my perspective as a user, I have done a lot of shopping over the last year or two, buying VPSs for clients. More recently I have been doing some research for a project of my own, and got bogged down in serious analysis paralysis I spent far too much time researching instead of doing.

  1. Some things are obvious - feeling safe with a big name; businesses being more willing to pay someone they have heard of.
  2. Another, more subtle, is that they have a lot of market share, and already have users familiar with their systems who find it easier to stick with what they know. That means these people are likely to stick with what they know for new projects.
  3. It is very easy to be confident that a large supplier will stay around. I rarely know how financially stable even fairly big businesses are (unless they are listed) let alone small VPS providers. What is they go bust tomorrow? What happens if the founder dies tomorrow?
  4. Even if they stay around, they may change. Will they scale up their infrastructure as they grow? Will customer service deteriorate? Will they be destroyed or severly disrupted by a DDOS? All have happened often enough.
  5. It is very easy to get information about big suppliers. AWS benefits a lot from this. The web is full of AWS specific how-tos, people recounting their experience of running almost anything you want on AWS and advice on how to size and tune for specific uses. AWS may perform worse, but you have a good idea of what you will get before buying. Most of all you have the reassurance of someone doing roughly what you are doing and getting it to work reasonably well.
  6. A lot of people commit themselves quite heavily to proprietary cloud platforms of various types before realising how much it will cost them, or how it will limit them. You then often have to rewrite stuff before moving which is not worth the cost....
  7. The big suppliers were ahead on "cloud" stuff like snapshots, quick scaling, and paying for what you use. A lot of people are happy to choose to probably pay more, if they know they will not pay for what they do not use. This is particularly true if usage is linked to revenues.
  8. A lot of suppliers have mixed reviews. Sometimes you can tell that it is the result of people not understanding what they are buying ("they would not help me install stuff on my unmanaged VPS") but sometimes it can be hard to tell. Again, this pushes buyers to safe options.
  9. It can be very hard to show that a smaller outfit is good. I always have doubts about whether performance measures are meaningful and relevant. It is easier to stick with "big brand and adequate" than "small and really good".
 
Last edited by a moderator:

drmike

100% Tier-1 Gogent
Tossing in my qualified opinion on why customers choose big companies for hosting and with emphasis on also paying higher rates.


1. Name recognition matters.  Real companies spend fortunes on establishing name recognition.  When you achieve commonly known status, the money is auto-pilot and just shows up.  Very little is needed to maintain healthy momentum and sustained business.


     1.A.  Look at brands like say GoDaddy with their Super Bowl ads.  Yes, those cost a ton, but GoDaddy has rode that wave for years.  I bet your mother knows who GoDaddy is, but has no clue who Digital Ocean is.


      1.B. McDonald's doesn't and hasn't historically done regular advertising.  Sure they spend on ads and always have, but their efforts are to remind you they are still everywhere.  As if those arches big and bright somehow lost visibility.  They do more of a supplemental advertising and on somewhat predictable schedules to up buyers during known cycles, mostly down cycles.


2. Brand is name recognition.  The more known a brand is, usually the more valuable / more dollars of transactions they are doing.   There is usually a strong relationship between the two.   There are exceptions notably in hosting where sites like LET have local celebrity brands ran by guys ever on the attention getting emphasis.  Those brands are potluck vs. the average stable and known company.  They may or may not be outperforming the stable staple companies.


3.  Trusting a company is going to be around is unsound.  This applies to hosting as much as it does to traditional consumer goods.  Technology however is a horrible industry where brands change hands all the time and the buyer axes the old company, murders the products, and mistreats the customers.  


    3.A. See Endurance International Group and their many botched deals from customer side. Arvixe comes foremost to mind.


4. The scale-on-demand and billed-on-use models are face value when new, appealing.  I tried these with Digital Ocean and ended up spending quite a bit more per month.   Reason why for me?  I don't have a stringed and tied deployment / control system.  Things I do are manual and no API dev at this time my side.  So resources were used for sandboxes and often teardowns of those didn't happen until end of month or days later, resulting in out of bounds spending.


    4.A. Flex billing and flex scaling are high end features for only the most advanced folks.  It's very niche.  With this comes higher expectations, SLAs, etc.  Those all cost to meet, require smart staff with experience, demands better support agents.  Read:  Cost centers galore. 


    4.B. Smaller companies can do flex things.  However, customers are going to pound them about the terms of service, demand SLA, push them to ramp up staffing.  All of these can kill even a well established smaller provider.


    4.C. Scaling is truly an enterprise level feature.  Sure some tech wunderkinds who like that stuff and engineer nifty thoughtful software to survive mass fails. The big industry though are rather big businesses.   Big business tends not to deal with small companies because small companies don't have various certifications on their DC builds, privacy, etc.  To a big biz, doing business with such a small company likely violates risk assessment and set policies established to keep the corporation safe and out of the Courts.


5. Most purchasing is done by peer group / endorsements by people you know.  Hosting certainly does the same for many.  Even here we recommend brands and there is group-think buy-in and try.  Small companies tend not to be in said circles with their company and products known / visible.   However Amazon and Google and other biggies are in there via proper channels of awareness they developed and marketed to.


6. Reviews are mostly rubbish for hosting.   


     6.A. There are paid blogger / affiliate sites hawking everything posing as content.  Blah.


     6.B. There are paid advertising sites meant to look like reviews. Blah.


     6.C. There are various forums where some legit customers share.  Finding content in such is -- blah.


     6.D. Reviews work when person reviewing something has tested something familiar to your own stack / needs / interests.  These aren't what you usually find when searching for company reviews.


7. Big companies on average probably perform worse than smaller VPS / cloud companies. 


     7.A. Tools and stack for most part is pretty common / similar.


     7.B. There aren't many enterprise magic tools out there.   Sure other virtualization in the paid world, sure some custom panels... But most of this stuff comes down to common gear, common CPUs, common loading, some best practices and some management layer to keep things from breaking ugly or proactively monitoring it all and API to the popcorn maker for when it all goes BOOM!


     7.C. There is no transparency or useful metrics across the industry.  Best we have is dd's and some peaking at values leaked and throughput tests plus IO wait.  These metrics all are flawed and subject to change at any time and might exist at level due to shaping, which is subject to change also.


     7.D. Business runs on money and profit.  It doesn't make bigger companies have more integrity or honesty about what they are selling.  Same corner cutting shenanigans are common in most hosting companies.  Corner cutting only happens where companies can truly hide things, points to transparency and the need for such as an industry standard with means to audit such.


7.5 Snapshots, quick scaling --- FEATURES.   Features may matter on something like AWS.  Locations might matter just as much.  Their certifications plus those features is likely the mix for the real money.


Feature driven companies are niche.  AWS is a quirky beast all said.  Highly complex, not approachable or very polished front side as a n00b.  It's a platform guys buy into because their counterparts in other shops did so.  Sure AWS works.   Performance per dollar metrics of that vs. a good independent VPS provider, pfft AWS loses. But indy provider lacks the certifications and SLA and yes, some features.  Indy company can certainly take customers from AWS, Google, or any of them and give a good cost savings.  Problem is owners of these indy companies aren't up to the marketing invest, spending to accomplish this or the time it takes to get momentum.


8. FAQ, Wiki, Knowledge base, etc. are all content if done well.  They are a form of lead generation if real content is developed that people search and find.  Companies like Linode and Digital Ocean have spent a ton on developing these and both are doing well in part due to that investment.


If a brand has this sort of data developed it certainly gives human / buyer signal that the company is legit and more than just a sales organization with the tired arse money grab.


Any brand can fake a website with content, design and images that look and sound 'professional'.  Having more content though, real support stuff, How-to's, etc. meh, I can't name one shitco that has such to any healthy level and if one does, odds are it's stolen or respun content.


So I am all over the place... Here's the deal in closing why customers choose big providers.  Big businesses globally are just trusted moreso.  To achieve that sort of success to be a bigger business, the customer logically thinks the company must be good enough or they never would have gotten that far.  All the above apply.  Name recognition is the butter on that bread for buying decision time.
 

graeme

Active Member
The reasoning is not necessarily correct, in fact I agree much of it.


Yes, big companies may not remain around, but it is usually easy to check on their finances. While brands change hands and service may deteriorate, they are unlikely to get shut down abruptly. They are not necessarily more stable, but I can, at least, evaluate their financial stability.


There is a lot of information about how well things work on big suppliers. Take a real example: a Google search I did about optimising Solr performance (a general one, not AWS specific) returned at least one blog post about optimising on AWS. A specific search gets lots of results. I doubt AWS is a particularly cost effective or good platform for Solr, but people have got it to work well, and there are instructions on how to do it.


Some reviews are useful, but I tend to find negative reviews more useful. Some complaints I find in reviews are clearly the buyers fault. Others are pretty credible (e.g. lots of people complaining about consistently slow customer service responses). A lot are in between.

My customers are small business, but willing to invest in fairly complex websites and webapps. In one case the business owner was familiar with AWS and wanted to stay with it. Not because of the brand, but because he could manage a lot of stuff himself without learning anything new. Another chose S3 rather than a dedicated server for large downloads even though a dedicated server would probably work out a lot cheaper because of the appeal of only paying for what was actually used. S3 was also more work for me!

I go out of my way to look for smaller suppliers for my customers and myself. The problem is the lack of meaningful reviews or metrics, which means I am limited to the few I know something about by reputation.
 
  • Like
Reactions: fm7

DomainBop

Dormant VPSB Pathogen
Why people pay so much for big brand VPS and cloud

One of the reasons is that a large percentage of small VPS providers (and even more so with small shared hosting companies) have limited technical knowledge and even less knowledge about the basics of running any aspect of a business:  cluelessness abounds in knowledge of legal aspects, finance, marketing, customer service (cover your ears anyone who is easily offended: but go read a few threads on LET where small hosts are exhibiting their lack of customer-centricness and you'll understand why 95% of these hosts will never make their living from being a host)


Typical statement indicating cluelessness and lack of business acumen you'll read from a small host on a VPS forum:

I have a friend whom i am helping with hosting. What info is a must keep for such? He thinks anything should just be handled pay pal side with out keeping anything but reference numbers and emails. He does not want to keep Names, address or invoices.

DrMike said (IPBoard's quoting is screwed up tonight):

Big business tends not to deal with small companies because small companies don't have various certifications on their DC builds, privacy, etc.  To a big biz, doing business with such a small company likely violates risk assessment and set policies established to keep the corporation safe and out of the Courts.

It's not only large businesses who are concerned with certifications and privacy.  Regardless of a company's size, if you're entrusting your company's data (which increasingly includes sensitive employee data and customer data as businesses move to "the cloud") to a provider/datacenter you need to be sure they can be trusted and certifications indicate the company's business has been audited by a 3rd party or the company was able to demonstrate its business practices meet certain standards.  Not doing your homework could be a costly legal mistake if your customer or employee data leaks because you entrusted it UnderageVPS.   There's a reason I've inserted this link in posts many times: Questions to ask your cloud provider

  1. Even if they stay around, they may change. Will they scale up their infrastructure as they grow? Will customer service deteriorate? Will they be destroyed or severly disrupted by a DDOS? All have happened often enough.

    In technology an equally important question is if will they adopt quickly enough to changing technology/a changing marketplace.  Let's pick on low end VPS providers again for an example: the low end market has been decimated over the past 2+ years (lots of failed businesses, closings, a ton of consolidation).  One of the reasons for the failure of many hosts is of course their unsustainable business models (or a complete lack of a real business plan), but the other equally important reason is the majority of businesses (most are not registered companies so I refuse to refer to them as companies) in the sector either haven't adopted to changing customer demand/changing marketplace, or have been too slow to adopt (yes, I'm talking about all of those cookie cutter SolusVM outfits who are late to the game in offering things like snapshots, hourly billing, etc, etc and now are either trying to play catchup or even worse are praying/waiting until SolusVM adds feature xx because no one in their business has the skills to implement feature xx and their finances rule out hiring someone to do it)
A lot of people commit themselves quite heavily to proprietary cloud platforms of various types before realising how much it will cost them, or how it will limit them.

Businesses need to do a thorough cost analysis before making any major commitment... (large corporate IT departments are just as guilty of this...going with solution A based on a slick sales presentation without really understanding what they just committed to)
 
Last edited by a moderator:

drmike

100% Tier-1 Gogent
I'll say to @DomainBop I won't buy from many companies, including the very largest because of lack of privacy and/or history of security fubars / disclosures.


The entire data backup schema out there drives me batty about handling and access if a company has such.  Knowing my data could be out there for well?  Ever perhaps?


People buy into big steadies like Amazon.  Problem I have there is I don't believe Amazon is running a common API, but proprietary everything.  A place like that scares me away as I am not into vendor lock-in.  I see whole companies running on AWS now.   It's one bad month for them from being in serious trouble and back to engineering in an ugly way.


Like I've said a lot lately, I prefer these days to self host or run dedis or colo gear just for sanity sake.  It's not perfect either.  But I remove layers of humans potentially up to no good.  Rogue employees borrowing things in those places.  Random extra dimensional fails with no one home.  It forces me to learn and document more on my end and engineer / paste together some more interesting solutions probably more appealing to a larger audience.


Heck AWS might be terrific, but I don't subscribe to mass adoption and popularity.... I just don't trust that company.
 
Last edited by a moderator:

graeme

Active Member
I prefer these days to self host or run dedis or colo gear just for sanity sake

What do you mean by self host other than running a dedi? Do you mean "self host and run dedis or colo gear"?

One of the reasons is that a large percentage of small VPS providers (and even more so with small shared hosting companies) have limited technical knowledge and even less knowledge about the basics of running any aspect of a business

Even if you exclude them from the discussion, a lot of what I saw applies to properly run businesses, not just the smallest and most clueless ones.
 

drmike

100% Tier-1 Gogent
What do you mean by self host other than running a dedi? Do you mean "self host and run dedis or colo gear"?

Literally what I said...


"I prefer these days to self host or run dedis or colo gear just for sanity sake"


I buy little virtual resource hosting anymore. If I need a slice, I cut one on a box I have, or spin up a new device for that.


I prefer to house things internally - on-lan --- which includes dedis remotely in some instances.
 

graeme

Active Member
Using my LAN is not an option for me (work from home freelancer). I have no existing boxes of my own.


I do not have any existing dedis, on the other hand I have two VPSs and two shared hosting accounts (each hosting multiple sites) just for myself, and I my clients have a lot more (including dedis where necessary).

I was thinking of consolidating all my stuff and some client stuff on a dedi. Even a low end dedi could replace a lot of small VPSs and shared hosting accounts and probably perform better. So far I am thinking on @drmike's lines


Then I changed my mind and thought a resizeable VPS from a good provider would be better. I read some threads suggesting that VPSs could perform better (faster disks for example) if not too heavily loaded, and, given that my (admittedly naive so far) testing suggests VPS can perform quite well and the sites I have on VPSs perform well, I decided that what I wanted was a good VPS, with enough scalability to meet foreseeable needs (not massively scalable either) and I was happily researching VPS (and cloud) suppliers. That experience triggered this this thread (triggered, but not the subject of, the post is mostly about other experiences over the last two years).

Now, @drmike's arguments in favour of using your own hardware has thrown enough doubt in my tiny mind to make me think again. Analysis paralysis worsening.
 
Last edited by a moderator:

graeme

Active Member
I am not sure that actually helps with the underlying problem: how do you find out whether small suppliers are 1) stable and 2) provider good networks, performance etc. short of the impractical and expensive step of trying a realistic test of each one that meets basic criteria, there is no way.


The big providers may well be worse on average, but it is a lot easier to get information. The market suffers from an information failure with regard to smaller suppliers. There is not sharp cut-off: for most requirements, I can definitely find some write up of experiences people with similar requirements on AWS, possible on Linode or DO, and almost certainly not on anyone smaller.
 

drmike

100% Tier-1 Gogent
Using my LAN is not an option for me (work from home freelancer). I have no existing boxes of my own.


I do not have any existing dedis, on the other hand I have two VPSs and two shared hosting accounts (each hosting multiple sites) just for myself, and I my clients have a lot more (including dedis where necessary).

I was thinking of consolidating all my stuff and some client stuff on a dedi. Even a low end dedi could replace a lot of small VPSs and shared hosting accounts and probably perform better. So far I am thinking on @drmike's lines


Then I changed my mind and thought a resizeable VPS from a good provider would be better. I read some threads suggesting that VPSs could perform better (faster disks for example) if not too heavily loaded, and, given that my (admittedly naive so far) testing suggests VPS can perform quite well and the sites I have on VPSs perform well, I decided that what I wanted was a good VPS, with enough scalability to meet foreseeable needs (not massively scalable either) and I was happily researching VPS (and cloud) suppliers. That experience triggered this this thread (triggered, but not the subject of, the post is mostly about other experiences over the last two years).

Now, @drmike's arguments in favour of using your own hardware has thrown enough doubt in my tiny mind to make me think again. Analysis paralysis worsening.

Home LAN is fine and always an option.  I have ample stuff hosted from the home office.  That stuff isn't client must have 100% uptime but hobby stuff, my own solutions, and development.  I could likely do client stuff and normal hosting just fine and no one would be any wiser.  In fact I did such for many years with some very busy websites serving millions of users a month.  This sort of hosting only becomes a problem where you are offgrid and lack power to run 24/7 or where your internet uplink is that bad.  For everything else, it's fairly doable, contrary to those deadset on everything in a datacenter (mainly because that's their bread and butter).


As far as boxes go, I use a LOT of ARM devices.  They are more than ample for lots of things.  Would I run WordPress on them?  Probably not, but that's more a WP issue than ability to run it. ARM devices are so cheap to acquire and so cheap to power each month.  Something like a quad core Raspberry Pi will run you like at most 50-75 cents a month in power. Meaning, it's economical as it gets.


I still use VPS containers out in the wild.  I use them for what I think I should be - remote tunnels, VPN, GRE, similar.  They are disposable in nature like biz model of many of the providers and their whims to suddenly terms slap you and rm -rf your data.


VPS where a company isn't massively oversubscribed on a node and where proper limits are in place on their end and RAID-10 actually is being used and perhaps cache too will trounce cheap alternative solutions your skimp to host your own.  But this is limited to disk and any time resource contention happens your container will suffer.  


All other VPS aspects are even more shared and subject to blips and slowdowns and therefore I'd say inferior.


I've wasted a lot of time cycles with a lot of providers over the years debugging lousy performance.  I've heard all the marketing sales speak and heard all the bogus stories on why things are working right, when they were not.


Time is a precious commodity no one has figured out how to produce more of.  Time > cheap outsourcing always.  This is why I've stepped back from the easy trap of virtual services and continue to lean towards DIY hosting.
 
Last edited by a moderator:

drmike

100% Tier-1 Gogent
I am not sure that actually helps with the underlying problem: how do you find out whether small suppliers are 1) stable and 2) provider good networks, performance etc. short of the impractical and expensive step of trying a realistic test of each one that meets basic criteria, there is no way.


The big providers may well be worse on average, but it is a lot easier to get information. The market suffers from an information failure with regard to smaller suppliers. There is not sharp cut-off: for most requirements, I can definitely find some write up of experiences people with similar requirements on AWS, possible on Linode or DO, and almost certainly not on anyone smaller.

Stability - this is always subject to change.  I'd be more interested with providers establishing their staffing numbers, and determining if anyone in ownership actually has technical ability.  Many shops are a mix of random people and technical folks too often are not the owners.  Meaning just one personal issue between that owner and that worker from the shop going technically stupid and stability through the floor.


Performance - performance points back to stability.  They tend to relate unless you are fortunate enough to be on an empty node.


Network is subject to change at any time.  Only way that can be controlled is to deal with companies that have their own blended networks and bandwidth commits.  Otherwise, most hosting companies get whatever their upstream / datacenter is giving them.  Bandwidth mixes change all the time and wreck DCs.  It's a constant sore spot for a lot of customers who care about performance, routing, latency, etc.


The market suffers from an information failure with regard to smaller suppliers.  


^^^ THIS.   I say this to clients / customers in different ways routinely.  I can explain the shortcoming too well.


Small shops lack robust team with proper planning. While some shops do well at technical, most aren't covering technical + sales + marketing + business planning + ???.  They suffer from one-dimensional approach to things where ownership is running all aspects and literally is mentally tired and overwhelmed.  It's hard to juggle multiple anything, with different business proficiencies to cover bases, well, those don't mix unless multiple people at happy hour.


They might want to develop a customer-centric culture as owners and point to some marquee hosting companies as 'did it right' namely on content to build customer-centric, but they can't get that done, dev own niche similar or push envelope much with their stressed resources.


When the content isn't the focus and customer centric is left to barely a helpdesk when things are failed, other customer-centric things won't happen.  So customers won't be out there with comprehensive reviews, there won't be true fanperson push from honest customers (there will be affiliates and folks on the money trip though muddying up the waters).


Sites otherwise won't get it together to flesh out the hosting industry.  Reason why?  It's work and the provider pool changes so rapidly that work done today at great expense is tomorrow invalid.  Even if those companies exist, so many moving pieces change that it just isn't worth it.   As a result, you are left with the industry shoe-horning information in unstructured forums and data derived from the best sales pushes each company makes.  Often that data is suspect at best and inaccurate at worst.
 

SkyNetHosting

New Member
Well I was wondering this when I was in WHD, I think its simply because cloud is the future and big companies want you to use cloud and they fund most of the bigger web hosting events etc. For bigger companies cloud is cheaper to deploy and they can make a better profit out of it.    
 

VpsAG

New Member
Hosting is not really a domain where most people like to experiment. The power user yes, but the average person would like something stable and well known which takes less of his time to manage.
 
  • Like
Reactions: fm7
Top
amuck-landowner