Why Are There So Many Damn Manufacturers?
Investors are always bewildered by the fragmentation in the security/surveillance market. They cannot understand why there are so many manufacturers and how none of them has ever been able to become dominant (i.e., 40 / 50 / 60 / 70 percent share).
At some level, it makes sense. There is just not enough product differentiation to justify the vast number of manufacturers in surveillance.
I have a new theory that I wanted to throw out.
Perhaps the reason for manufacturer fragmentation is because of the huge number of integrators. Integrators tend to purchase differently than end users. Instead of just worrying about what is the best fit for their personal needs, they need to anticipate what products will help them sell the most / best.
Integrator Motivation For A 'Unique' offering
A regular theme with many (but not all) integrators is having a product that your rivals do not (whether it's a line that is dealer protected or one that is just not widely offered in your market). Whether that line is objectively better or worse, leading with a line your rivals do not offer, gives the integrator some sort of uniqueness that can then be sold / marketed as a differentiator.
Likewise, I suspect that until the number of integrators decrease and / or end users start buying direct without integrators, the market will continue to support hundreds, if not thousands of manufacturers.
What do you think?
I agree. I also believe Integrators are obviously looking for the best margin and try to differentiate from competition with some original solution, especially for small systems (<16 IP cameras)
Leading manufacturers provide most of the time less than 20% gross margin (during quotations difficult to keep more than 5%, like for Cisco …..) All prices can be compared on the Net, so every time Integrators can find a new small “me too” brand they try to integrate it or…to source and brand it when they have some sales potential. It does explain in Europe we have more than 20 purely artificial flavor brands which are sub productions from some Tw , Cn or Kr manufacturers met during an international Cctv exhibition. Obviously till Onvif 2.0 isn’t well integrated everywhere it causes major installation, Integration and maintenance issues especially when integrated with other systems.
What’s funny is that with Onvif 2.0 and latest version , you will be able to use more “unknown” brands than before, because they will be recognized by major Vms editors as major cam brands are.
Well if you look at all the offerings out there none seem to be "perfect". To dominate you got to have the "holy grail" (e.g. Intel has powerful AND efficient CPUs).
Cameras: Axis has WDR and Lightfinder but not yet a holy-grail camera that has high 2MP+ resolution with automatic switching between WDR and low light boosting modes, intelligent IR for complete darkness, and availability in bullet and dome formats.
Analytics: VideoIQ is great for security but don't sell their software separately forcing you to use their NVRs or cameras. Also they do not have a retail counting/business intelligence offering. On the other hand 3VR and AgentVI do not do full resolution 1MP+ analytics processing so they have limited range.
VMS: There are tradeoffs EVERYWHERE: Exacq is efficent but the playback/investigation is clunky and there is no software based motion detection. Milestone has issues that can crash the server if the camera supplies "malformed" video streams (this is an architectural issue - to do with the fact that the recording server is tightly couple to DLLs for its camera drivers and the server crashes when the DLL crashes), Luxriot has no mobile app and does not support granular storage management, Genetec is too complex for SMB users, HD Witness has compatibility issues with older GPUs and certain cameras, NUUO is very inefficient but has good audio support, Smartvue has excellent ease of use and bookmarks/comments (rare feature) but very poor multi-grid live view support (4fps). There really isn't a go to VMS that is balanced.
PoE Switches: Generic brands do not have options that scale to enterprise deployments (L3+ management) while Cisco does not have any real high power low end unmanaged offerings.
Storage: Generic brands do not have options that scale to enterprise deployments (SANs with redundant PSUs and automatic failover) while mainstream players like Dell/HP does not have any real entry level offerings for 3-8 drive setups.
Back in 1981 I visited the Akihabara district in Tokyo. I was absolutely dumbfounded by the diversity in product offerings. Where I had seen a few models of microwave ovens in the US (for instance), there were literally hundreds of models on display. So why the profusion? It seems inbuilt to the engineering spirit to try almost endless variations on a theme.
No offering is perfect and no standard truly complete. A previous commenter postulated that even weak standards fertilize smaller competitors.
All of the previous commenters had good specific points but I think this "search for the holy grail" is a deep-rooted almost genetic drive that flowers (particularly with creative types) in relatively free societies and is not really subject to details of market opportunities.
Bohan, Paul, technical differences can drive a certain number of competitors but 500 manufacturers, 1000 of them? Are we really saying that anywhere close to all of these competitors is a product of pure technical differentiation? What percentage of manufacturers have significant technical differentiation against rivals? If you walked the floor of an international security show (including the back), how many?
Let's look at the market share of other industries. Here's DSLR:
And here's IP Phones:
And here's search engines:
Most markets are like thisand that's why investors are bewildered about surveillance. Insurveillance, the top 10 combined are equal to the number 1 in most other markets. Is surveillance so much more sophisticated than any other market? Is the drive for technical perfection much greater? Or is there another factor?
John , you cant' compare very high tech technologies, I mean Core switches for operators requiring very high level trained engineers with smal 100 to 1000 camera euro devices. The Market entry level cost for bowser is very high, but it's very easy I presume to launch a new unknown name from Chinea or Korea and say I'm 30% cheaper but I'm also compatible with Milestone, Genetec... If tomorrow you dont' care your camera brand because all brands react the same way on a technical point of view ....the market will remain fragmented.
But The metoo brands without any "Plus " product and not price competitive will surely die because there is no place for it, (I have few brands in mind..) as you need a mid term maintenance policy that some new players can't offer from China without local infrastructures, inventory and investments. This local presence costs and eliminate too small brands.
When you do large video surveillance installations you need to provide anticipated RMA procedures, sometimes in direct, not through your distri partners to shorten delays (SOny, Axis are doing it) . That's a key argument for most large accounts. According to me, it will kill the small and too new players when the market will become more mature, which is probably next 5 years milestones: 100% IP , Onvif 2.0, getting more trained and aware of potential issues, improve preventive maintenance ...
Surveillance is a 'fringe' product for most of the world. Unlike search engines or soda pop, where the market is consolidated because of brand exposure and refined personal taste, surveillance equipment is still the 'wild west'.
Is that due toindustry sophistication? Heck no. It's because in the jumbled up and busy head of most end users, they simply haven't had the chance to learn who the 'big' names are, so they take what they frantically search off the internet or what brand that 'strange electronics/computer dude' deep in the Rolodex recommends.
We (IPVM membership) have a tilted viewpoint on this, because:
a) We're better informed than most about the market/positioning/products
b) We CARE to know (either because of obligation or interest), and it is important to us that we do.
Talk to 'average Joe' small businessman, and he knows nothing about video surveillance beyond needing cameras and some kind of recorder (maybe). Ask that same guy 'iPhone or Android?' and you're likely to get an earful. Personal preference is king in surveillance, because there is minimal personal risk in expressing it professionally.
I hate to say this, but maybe the security marketers suck. They devote so much effort to such a small audience, the larger population is missed.
Maybe Avigilon isn't so crazy advertising in WIRED magazine, after all:
Avigilon was not crazy at all. How else would a subpar system like Cisco Surveillance get into the market?
Any other theories out there on why the surveillance market is so fragmented?
I would second Bohan's statement "Well if you look at all the offerings out there none seem to be "perfect"." There's not one manufacturer I can think if that does enough, well enough, to be a sole product solution. It's like they've conspired together to divide up desirable aspects without any particular one being too strong to be overwhelming.
The security market appears to be undergoing a technology refresh with the whole IP thing in the last few years. This reminds me a lot of the early PC days in the mid-80's. You had tons of companies competing, lots of "Store Brands", multiple OEM's of the same thing sold under different names and so forth. If anyone remembers "Computer Shopper" it was a monthly catalog of hundreds of pages of nothing but PC gear, now I'm not even sure it's still published.
I think over time we'll see some big names emerge, and a lot of smaller ones drop out as things standardize around various brands, formats (resolution/aspect ratio), and feature sets.
I also agree with John's assement that many integrators try to differentiate themselves from their local competitors by carrying unique products. That will continue to prevent massive market share grabs and a consolidation down to just 3 or 4 main brands, but we'll see a lot of the non differentiated lines fade away.
The role of end users is important here as well. It's not simply that integrators are pushing lots of random and/or relabeled offerings, it's that end users are buying them.
What, if anything, will make end users push back? Will the Internet make them be more informed? Will IT buyers of surveillance be more selective than security / facility buyers of surveillance?
To the original question of why so many Manufacturers, I think it's because the barriers to entry are fairly low and companies thinks they may have a pre-made market with their existing customer base
In the example of Cisco, they assumed since they sold networking to so many customers, it would be a very easy to move to IP video relying on nothing but the Cisco name. GE did the same thing back in the 90s (Kalatel acquisition) and it was moderately successful.
Going back to my integrator days, I have another good example. A very significant end user I had in the 90's who was buying fire safes from a very good manufacturer of fire safes. Once DVRs started to replace VCRs this company decided to build DVRs. This one significant end user was enough to propel that business line forward. As a DVR the product made a great doorstop. But we sold and supported the product best we could, even though we strongly advised against it. Eventually it was replaced by the end user but not until thousands of the units had rolled out
I love this subject, it's one I expound on regularly and have a few ideas but no concrete answers (sorry!). Firstly there appears to be no real barriers to enter this market as a manufacturer - as long as you pass CE/FCC you can make a camera and sell it globally, so that's attractive to a couple of thousand guys in Shenzen... and everyone that then OEMs it. Secondly I think there is a level of ignorance about the value of the market globally, OK there are numbers published but how accurate are they? Is this a $14Billion opportunity or $40Billion and is that an end user valuation including install etc? Then there is optimism about the achievable market share as mentioned previously - who sets out to achieve a 0.01% market share?
The technology sector comparison I sometimes use is business telephone systems - nearly every SME upwards has one but how many manufacturers are there? 20 -50 is my estimate and how much is that sector worth (164million IP telephones sold in 2010 for example), is the technology so advanced that only a few companies can make phones, are the end users driving product development? I don't think so. Here too we have installers and integrators, how do they differentiate when all their competitors have access to the same equipment - service and price as usual.
It amazes me how competitive our industry is; how massively contested it is. Life might be easier if we didn't have to compete with so many other developers, manufacturers and OEM companies, but it is driving the technology forward faster and making it more affordable, so the End User is gaining in that respect. Will it change any time soon? No.
John H, glad you liked that analogy from the mid late 90s.
I'm sure you and the readers can remember that was really the explosion of DVRs both good and bad. In our company we were seeing new boxes from manufacturers we'd never heard of coming in almost monthly. Many were "found" by end users who were sure they were getting a great deal. Just like now price was a driving factor when they compared it to a stable product we'd recommend like Katatel or DM. I dreaded when I heard my customers were going to ISC West because I knew they'd come back with something horrible we'd have to install.
Some of the DVRs seemed like they were built in a garage. All you needed was a tower PC & video capture cards and voila - you were a DVR manufacturer. We joked, the life expectancy of one called a Dana was 6 months and 1 day. That's because it came with a 6 month warranty.
In some ways were in the same sort of era now. People will hire a code writer and "boom" they're a VMS. Or they'll source a Chinese cam and stick it in a housing they have built to spec and now they're a camera manufacturer.
Yeah, why can’t people just stick with a tried and true like Polaroid.
For an industry to consolidate, there has to be perceived economy of scale to make the consolidation attractive from a profit standpoint. In other words, one company with a large enough sales quantity can lower their margins so that they make less per unit (increasing competitiveness) but sell enough units to make up the difference.
I don't see that happening in this industry yet because it's too fragmented. But I'm pretty sure it will happen eventually.
Mobotix financials review: "High Profit Margins" "Their profit margin is up to 14%..." "Gross profit increased 2.6 points from 69.3% to 71.9%"
By comparison, (for example) Walmart's recent gross/net profit margins were on the order of 25%/3.6%.
Video surveillance is substantially more complex and challenging to integrate than a toaster or a banana. Compared to surveillance, Walmart's market is much closer to the theoretical "perfect market" with commodities that are interchangeable, lowering barriers to entry and driving down profit margins. Much of Walmart's success, and a standing barrier to new entrants, has been as much their very low margins enabled by substantial levels of vertical integration and by volume purchasing. How is (for example) Amazon taking a piece of that pie? Even lower margins, on the order of 1% net after taxes.
IPVM is a testament to the complexity and lack of commoditization of video surveillance. Many many users are asking simple to complex questions about challenges and expectations with various pairings or, indeed, even with very "intuitive" implementations.
So, with sweet margins and enough complexity to lock in the captives, what can possibly force a consolidation? It reminds me of the early days of computing with the Osborne, Kay, Apple, Commodore, Sinclair, Atari, and, oh yes, early IBM computers. None terribly compatible. Fairly complex. With tools like Visicalc and WordStar, it was easy to understand the promise, but difficult to achieve it simply.
Somehow IBM became a de-facto standard, and they effectively created Microsoft when they outsourced the operating system. Even in the heyday of IBM PCs, I recall going back and forth from DOS 5 and 6 to Windows 3.0 several times - the promise of common user interface and clipboard data exchange was compelling but the obstacles, immaturity, and crashes still daunting. Remember "plug and play?" That alone must have taken 5 years to refine, with the whole journey from IBM PC/PC JR to grandma emailing the grandkids probably taking about 15 years.
Computer "killer apps" varied from early mostly business uses (spreadsheet and word processor) to rapid communications (early internet, both business and personal email) to a huge knowledge base that almost everyone uses every day.
New adopters for these "killer apps" were the real ground swell that moved the market forward, since many die-hard legacy owners struggled onward for a time.
I wonder what might be likely killer apps within video surveillance or security? Something that meets a price/performance that every home will want?
Who could be a likely "IBM" of the video surveillance or security world today?
Do you see current users abandoning their Timex Sinclar or Atari some other not yet clear "global" standard? (VHS vs Betamax comes to mind)
Can you comment a bit about HD-SDI. I woukd have thought it would do well in the Casino vertical due to low latency and optimal live view quality?
Neither HD-SDI nor HDcctv are well-suited to large-scale systems, in my opinion. The approximate 100 meter transport limitation over runs of coax and lack of current UTP support severely restricts how many cameras we could convert (probably less than 40% of our installation). Mid-span extenders will not help this - we would have to find the midpoint of one cable buried among many in cable trays and find a place to install each extender.
Then you have system compatibility. There are few, if any, standalone encoders for HD-SDI / HDcctv and the ones that are available typically won't integrate with any major VMS. HD-SDI and HDcctv are, at least at this point, suitable only for small applications using DVRs.
Finally, the so-called advantages in regards to "live" video clarity and low latency are only available if the cameras are run through matrix switch(es) directly to monitors. I haven't seen any suitable matrix switches. What's available appears to be based on broadcast and studio applications and are extremely limited compared to modern large-scale analog matrices like the Pelco CM9700 series.
Video Surveillance Show Armed Suspect Trying To Enter Club, Security Intervenes, Fight Breaks Out (1)
Started by Jermaine Wilson
|1 minute by Jermaine Wilson|
Started by David Little
|2 minutes by Undisclosed Manufacturer #2|
Started by Undisclosed Integrator #1
|less than a minute by Undisclosed Integrator #1|
Started by John Honovich
|less than a minute by John Honovich|
Started by Brian Rhodes
|less than a minute by Brian Rhodes|
Back to Top