Forrester just came out with an interesting report “Should Your Email Live In The Cloud? A Comparative Cost Analysis” (You can get a free copy over at Network World)
Amongst other things the report found that the majority of executives at the target enterprise companies, when asked to estimate, believe their per email box costs where typically 4 times less than the actual costs turned out to be when using a Total Cost of Ownership evaluation approach.
The real interesting thing for me, however, was that the majority of firms asked said they would be pursing a hybrid on premise and cloud approach for their email needs:
Whilst the core mailboxes themselves may be based on CPE (Customer Premise Equipment) – elements of the overall email system would be outsources to one or more cloud providers – this included functionality such as mail filtering, spam filtering, storage and backup, DR and mobile mailboxes.
The insight here for MSP product mangers is that enterprises and SMBs’ rarely make the huge leap to out-tasking an IT function in one step – and the needed for graduated portfolio of offerings so that the end customer can move to a fully outsourced consumption model over time is key to a successful (and very adaptable) value proposition.
Co-location Providers learned this lesson with so called Co-location Plus offerings – with piecemeal offerings such as managed firewall, managed load balancing, DB management, OS management, shared and dedicated storage offers – all designed to penetrate further and further into that customers collocation cage until they look a lot more like a managed customer than a colo customer.
I think we will see the emergence of similar graduated portfolios in the Cloud Provider – addressing the potential compute requirements of the enterprise in bite sized pieces – whether that is burst capacity, on demand development / test environments, failover / DR compute or batch data possessing – the slowly slowly approach is needed to bring the enterprise and SMB to the benefits of out-tasking and cloud.
Tuesday, January 26, 2010
Friday, January 22, 2010
3 ways to innovate and differentiate managed and cloud services
In 2010, managed and cloud service providers will need to continue to innovate in this pre-chasm market with the development of new features and business models as they strive to differentiate themselves in a market that is already commoditizing.
With ongoing pressure to find new revenue streams, defend against competitors and develop deep and wide relationships with their end customers, the MSP Product Manager certainly will have a difficult equation to solve. And, unfortunately, the MSP is often selling to a market that continues to be very price conscious and expects the cloud promise of "compute for less". As a result, developing the "Cadillac" of cloud offerings, whilst perhaps commanding a premium price, will have a limited market and MSP Product Managers will need to look to others ways to innovate to differentiate.
1. Product Strategy: One approach that I am starting to see providers have success with is in developing offerings that adopt the principal of “Mass Customization”. Mass Customization sees an offering enhanced with a plethora of configuration options (dials and switches as it called) so that vertical markets and horizontal segments can be pursued with an offering that appears tailored to that market’s need. Within that framework, the end customer has the option to further customize the offering specific to their needs – further enhancing to the “built just for me” perception.
2. Service Delivery: However developing offerings bristling with dials and switches is not sufficient to “solve the equation” alone. Provisioning and delivery must be codified and industrialized through a combination of Service Orchestration, Service Provisioning and Service Management technologies on the back end. Then any combination of settings on the "dials and switches" can be fulfilled in a routinized manner and the service should be designed to achieve similar economies of scale to a “one size fits all” offering.
3. Go To Market: Furthermore these highly configurable services must be complimented by well thought out Go To Market (GTM) strategies: target verticals will often need to be approached via a channel they already trust and this evolution of offerings will inevitably see partnerships between scalable MSPs and vertical focused VARs, Sis and niche MSPs equipped with vertically resonate value propositions.
As the cloud services market continued to mature other innovation options will present themselves and become viable and appropriate - for now, however, the pressure is going to be on the product manager to innovate at a product feature level.
With ongoing pressure to find new revenue streams, defend against competitors and develop deep and wide relationships with their end customers, the MSP Product Manager certainly will have a difficult equation to solve. And, unfortunately, the MSP is often selling to a market that continues to be very price conscious and expects the cloud promise of "compute for less". As a result, developing the "Cadillac" of cloud offerings, whilst perhaps commanding a premium price, will have a limited market and MSP Product Managers will need to look to others ways to innovate to differentiate.
1. Product Strategy: One approach that I am starting to see providers have success with is in developing offerings that adopt the principal of “Mass Customization”. Mass Customization sees an offering enhanced with a plethora of configuration options (dials and switches as it called) so that vertical markets and horizontal segments can be pursued with an offering that appears tailored to that market’s need. Within that framework, the end customer has the option to further customize the offering specific to their needs – further enhancing to the “built just for me” perception.
2. Service Delivery: However developing offerings bristling with dials and switches is not sufficient to “solve the equation” alone. Provisioning and delivery must be codified and industrialized through a combination of Service Orchestration, Service Provisioning and Service Management technologies on the back end. Then any combination of settings on the "dials and switches" can be fulfilled in a routinized manner and the service should be designed to achieve similar economies of scale to a “one size fits all” offering.
3. Go To Market: Furthermore these highly configurable services must be complimented by well thought out Go To Market (GTM) strategies: target verticals will often need to be approached via a channel they already trust and this evolution of offerings will inevitably see partnerships between scalable MSPs and vertical focused VARs, Sis and niche MSPs equipped with vertically resonate value propositions.
As the cloud services market continued to mature other innovation options will present themselves and become viable and appropriate - for now, however, the pressure is going to be on the product manager to innovate at a product feature level.
Tuesday, January 19, 2010
CIO Attitudes to managed and cloud services
Following on my earlier post on the Cloud Hype cycle, a new report from Data Monitor (commissioned by BT) finds that 57% of CIOs are still not comfortable with running applications and storing corporate data on computing environments based outside their country. I wonder if this is just about geography or if it’s actually about computer environments outside of their direct control and domain.
There is no doubt we are on the Cloud Computing Hype-cycle, and we have some way to go, including a sojourn through the "trough of disillusionment", before we reach the "plateau of productivity". No doubt we have some way to go before this consumption model hits main street with many issues around security, standardization and reliability needing to be solved first. (note - if you get a chance, get "The Big Switch" by Nicolas Carr - this draws the interesting analogy between the Electricity Generation at the turn of the 20th century and today’s nascent cloud computing market)
We have seen this before with the Web Hosting market in the mid 90's. Many IT departments, with no viable alternative, implemented and managed their web infrastructures themselves. IT managers in SMB's were running T1's to their building and clustering a couple of Compaq servers to host their web / e-commerce presence. As the Web hosting market emerged and matured, these requirements were increasingly out tasked. Today - the web hosting market is as mature as it is commoditized, and out tasking this IT function is the norm rather than the exception. Inevitably, the same is going to happen for Cloud Computing and Storage offerings. However, as with web hosting, a level of natural offer and market maturation will need to take place before these services cross Moore’s Chasm into main street. MSP product managers need to build this reality into their hockey stick revenue assumptions. I am not saying that it won't be a "hockey stick" - we just need some realism on how soon the exponential upturn in revenue is likely to happen
There is no doubt we are on the Cloud Computing Hype-cycle, and we have some way to go, including a sojourn through the "trough of disillusionment", before we reach the "plateau of productivity". No doubt we have some way to go before this consumption model hits main street with many issues around security, standardization and reliability needing to be solved first. (note - if you get a chance, get "The Big Switch" by Nicolas Carr - this draws the interesting analogy between the Electricity Generation at the turn of the 20th century and today’s nascent cloud computing market)
We have seen this before with the Web Hosting market in the mid 90's. Many IT departments, with no viable alternative, implemented and managed their web infrastructures themselves. IT managers in SMB's were running T1's to their building and clustering a couple of Compaq servers to host their web / e-commerce presence. As the Web hosting market emerged and matured, these requirements were increasingly out tasked. Today - the web hosting market is as mature as it is commoditized, and out tasking this IT function is the norm rather than the exception. Inevitably, the same is going to happen for Cloud Computing and Storage offerings. However, as with web hosting, a level of natural offer and market maturation will need to take place before these services cross Moore’s Chasm into main street. MSP product managers need to build this reality into their hockey stick revenue assumptions. I am not saying that it won't be a "hockey stick" - we just need some realism on how soon the exponential upturn in revenue is likely to happen
Tuesday, January 12, 2010
Top 5 Mistakes in Managed and Cloud Services Business Models
For any product or service, being able to create realistic and achievable business models is key to setting expectations as to the revenue and profit potential. This particularly applies to managed and cloud services, which often have high up front investments and barriers to entry, as well aslong complex, sales cycles. Especially with managed and cloud services, careful consideration needs to be paid to the variables that affect the cost and revenue curves: they both have sensitivities that are not immediately apparent.
Throughout my career, I have built many managed and cloud services, and I also have reviewed many partners’ business models - I wanted to share with you the top 5 mistakes I have come across.
1. Customer On-Boarding Rate
An easy mistake to make - I have 100 customers billed at $100 dollars a month - what is my revenue for year 1 ? 100 x $100 x 12 = 120,000 right ? Wrong - consideration has to be made using the rate at which these customers come on board. If they come on at a linear rate over the course of the year (i.e. you exit the year with 100 customers), then to approximate the revenue you divide the above by 2 - so the revenue would be $60,000. More complex "curves" can be used to model the customer on-boarding rate.
An associated area, the so called "Rule of 78" (a term borrowed from the finance industry) also has to be considered. If you miss your new customer target in Month 1 by 5 - you have to get more than 5 additional new customers (on top of the existing target) in Month 2.
2. Churn
In some commoditized managed and cloud services markets which have relatively low switching costs for the end customer, consideration has to be given to the "Churn Rate", meaning how many customers will NOT renew after their initial contract rate. This can be low, single digit percentages - but can reach up to 15% in more cut throat markets.
3. Hockey Sticks
The classic hockey stick revenue curve. I launch my service day 1, and it just keeps getting exponentially better and better from there on out. Product Managers have to consider the lag between launch and First Customer Acquisition, they need to think about sales cycles, they need to think about provisioning timelines, bedding in and turn up periods. It’s not unusual to see no revenue coming into the business for 6+ months after the launch. This is the reality of the managed and cloud services business - particularity in cases where the organization is entering into new markets. Often, many aspects of the company have to be re-tooled: sales, marketing, customer relationship and finance. It doesn't matter how great the service offering is - if these pieces are not in place, then revenue goals will not be achieved.
4. Step Function Costs
Managed and Cloud Services have high barriers to entry - much higher than the costs for a VAR or even a Professional Services Systems Integrator. OSS Costs, BSS Costs, NMS Costs 1. Don't even get me started on OSS / BSS / NMS costs.
5. NOC Staff Costs
On the subject of Step Function Costs. Usually, the largest part of a 5 year cost profile is the NOC Staff Costs and Customer Engineer Costs - even for organizations that are already in the business of managed and cloud services they can be prohibitive. For example - if you want to have 2 people in your NOC 24 x 7, then you need to hire 12. That is just the way the math works when you think about training, vacation, work weeks, failover capacity, etc. 12 people at, let’s say $120K loaded annual salary – that’s $1.4M .... day 1 ... as you go live with customer 1. This is the area that I usually see the grossest underestimation - so working with your service delivery teams and organizations to make accurate estimates of this is key. However,this is not a blank check for your NOC manager - always be challenging the assumptions. What existing skills sets can be leveraged? What tools can be used to drive closure of tickets to your level 1 NOC teams? What technologies can be put in place (such as event correlation technologies) that reduce the noise that comes up to the NOC teams? What knowledge bases can be put in place to drive repeatable codified ticket closure ... all these aspects should be considered when modeling this key cost center.
I have collected a number of tools over the years that can be used for this complex task - if you are interested in getting your hands on some then drop me a note
Throughout my career, I have built many managed and cloud services, and I also have reviewed many partners’ business models - I wanted to share with you the top 5 mistakes I have come across.
1. Customer On-Boarding Rate
An easy mistake to make - I have 100 customers billed at $100 dollars a month - what is my revenue for year 1 ? 100 x $100 x 12 = 120,000 right ? Wrong - consideration has to be made using the rate at which these customers come on board. If they come on at a linear rate over the course of the year (i.e. you exit the year with 100 customers), then to approximate the revenue you divide the above by 2 - so the revenue would be $60,000. More complex "curves" can be used to model the customer on-boarding rate.
An associated area, the so called "Rule of 78" (a term borrowed from the finance industry) also has to be considered. If you miss your new customer target in Month 1 by 5 - you have to get more than 5 additional new customers (on top of the existing target) in Month 2.
2. Churn
In some commoditized managed and cloud services markets which have relatively low switching costs for the end customer, consideration has to be given to the "Churn Rate", meaning how many customers will NOT renew after their initial contract rate. This can be low, single digit percentages - but can reach up to 15% in more cut throat markets.
3. Hockey Sticks
The classic hockey stick revenue curve. I launch my service day 1, and it just keeps getting exponentially better and better from there on out. Product Managers have to consider the lag between launch and First Customer Acquisition, they need to think about sales cycles, they need to think about provisioning timelines, bedding in and turn up periods. It’s not unusual to see no revenue coming into the business for 6+ months after the launch. This is the reality of the managed and cloud services business - particularity in cases where the organization is entering into new markets. Often, many aspects of the company have to be re-tooled: sales, marketing, customer relationship and finance. It doesn't matter how great the service offering is - if these pieces are not in place, then revenue goals will not be achieved.
4. Step Function Costs
Managed and Cloud Services have high barriers to entry - much higher than the costs for a VAR or even a Professional Services Systems Integrator. OSS Costs, BSS Costs, NMS Costs 1. Don't even get me started on OSS / BSS / NMS costs.
5. NOC Staff Costs
On the subject of Step Function Costs. Usually, the largest part of a 5 year cost profile is the NOC Staff Costs and Customer Engineer Costs - even for organizations that are already in the business of managed and cloud services they can be prohibitive. For example - if you want to have 2 people in your NOC 24 x 7, then you need to hire 12. That is just the way the math works when you think about training, vacation, work weeks, failover capacity, etc. 12 people at, let’s say $120K loaded annual salary – that’s $1.4M .... day 1 ... as you go live with customer 1. This is the area that I usually see the grossest underestimation - so working with your service delivery teams and organizations to make accurate estimates of this is key. However,this is not a blank check for your NOC manager - always be challenging the assumptions. What existing skills sets can be leveraged? What tools can be used to drive closure of tickets to your level 1 NOC teams? What technologies can be put in place (such as event correlation technologies) that reduce the noise that comes up to the NOC teams? What knowledge bases can be put in place to drive repeatable codified ticket closure ... all these aspects should be considered when modeling this key cost center.
I have collected a number of tools over the years that can be used for this complex task - if you are interested in getting your hands on some then drop me a note
Monday, January 11, 2010
Turning Service Levels and Outages to your advantage
Another outage and blow for Cloud computing, today 68,000 Salesforce.com customers were without service for over 1hr.
This is on the back of an number of high profile outages in 2009 including a large outage from Google in May, Rackspace in June and Amazon amongst others. All the surveys and market intelligence I have seen will rank "Reliablity" in the top 3 as being a major concern and inhibitor for enterprise customers when considering managed and cloud services (and IT out-tasking in general) as a means of consuming their Information and Communication technologies.
The "I told you so" guys are always going to make hay out of these kind of events - to counter this we should ask (and train the sales force to do that same) What kind of reliability does the average enterprise or SMB experience with applications and hardware they run themselves ?
What should be highlighted in the Managed and Cloud Services value proposition - is not only the feature function of the service itself - but the investment the following.
People: Show the NOC staff, skill sets, certifications, tiering, knowledge bases, and training programs that you have invested in.
Process: Talk about the processes (ITIL), escalation, continuous performance, ISO and six sigma processes that are in use
Technology: Finally - show how the people and process are enabled by the investment in the provisioning, monitoring, service orchestration, NMS, Correlation, Performance Management and ticketing systems that has been made to support and service and stand behind the SLAs.
In my years on the road in front of customers - with almost no exception there was never a case where the custom had made anything like the investment that the mature MSP has. Couched correctly - you can't be arrogant about it - but a respectful and authoritative review of the above can be a real boost in the MSPs value proposition.
This is on the back of an number of high profile outages in 2009 including a large outage from Google in May, Rackspace in June and Amazon amongst others. All the surveys and market intelligence I have seen will rank "Reliablity" in the top 3 as being a major concern and inhibitor for enterprise customers when considering managed and cloud services (and IT out-tasking in general) as a means of consuming their Information and Communication technologies.
The "I told you so" guys are always going to make hay out of these kind of events - to counter this we should ask (and train the sales force to do that same) What kind of reliability does the average enterprise or SMB experience with applications and hardware they run themselves ?
What should be highlighted in the Managed and Cloud Services value proposition - is not only the feature function of the service itself - but the investment the following.
People: Show the NOC staff, skill sets, certifications, tiering, knowledge bases, and training programs that you have invested in.
Process: Talk about the processes (ITIL), escalation, continuous performance, ISO and six sigma processes that are in use
Technology: Finally - show how the people and process are enabled by the investment in the provisioning, monitoring, service orchestration, NMS, Correlation, Performance Management and ticketing systems that has been made to support and service and stand behind the SLAs.
In my years on the road in front of customers - with almost no exception there was never a case where the custom had made anything like the investment that the mature MSP has. Couched correctly - you can't be arrogant about it - but a respectful and authoritative review of the above can be a real boost in the MSPs value proposition.
Pricing Models: Amazon Spot Instances - a Sign of Things to Come ?
I saw that Amazon Web Services (AWS) just announced a new feature of their highly successful Cloud offerings (I say highly successful - at least AWS is the poster child for cloud computing right now - not sure how much money they are making off it)
However - they announced a new option for purchasing and consuming Amazon Elastic Compute Cloud (EC2) compute resourced call Spot Instances.
This feature enables customers to bid on excess unused Amazon EC2 capacity and use those instances for as long as the "bid" exceeds a derived dynamic Spot Price. This Spot Price is based on supply and demand and does change - when a customers bids exceed the spot price they gain access to available Spot Instances. This compliments existing purchase features on AWS EC2 - such as Compute Futures - where you can reserve future compute capacity.
My sense is that today these features are more of a gimmick than and legitimate way of purchasing - but it is is definitely the sign of things to come and a leading indicator of the ultimate and eventual commoditization of compute and storage resources - just like the commodiitzation of electricity at the turn of the 19th century (See Nicholas Carrs great "The Big Switch" for more on this analogy)
More here at Amazon AWS
However - they announced a new option for purchasing and consuming Amazon Elastic Compute Cloud (EC2) compute resourced call Spot Instances.
This feature enables customers to bid on excess unused Amazon EC2 capacity and use those instances for as long as the "bid" exceeds a derived dynamic Spot Price. This Spot Price is based on supply and demand and does change - when a customers bids exceed the spot price they gain access to available Spot Instances. This compliments existing purchase features on AWS EC2 - such as Compute Futures - where you can reserve future compute capacity.
My sense is that today these features are more of a gimmick than and legitimate way of purchasing - but it is is definitely the sign of things to come and a leading indicator of the ultimate and eventual commoditization of compute and storage resources - just like the commodiitzation of electricity at the turn of the 19th century (See Nicholas Carrs great "The Big Switch" for more on this analogy)
More here at Amazon AWS
Managed and Cloud Services help Cross the Chasm
I just finished Geoffrey Moores "Dealing with Darwin" - I have to say a really great book providing a very thorough way of thinking about different types of innovation available to a company as they products mature through the technology adoption lifecycle.
As I was looking at this lifecycle I was thinking how a enterprise or SMBs purchase criteria would change across that lifecycle as the product or services moved from appealing to Innovators and Early Adopters across the chasm to the Early and Late Majority (also known as Main street) and eventually the Laggards. Each type of organization will have different criteria when selecting technologies and therefore vendors need to create different value propositions depending on where their product or service is in the lifecycle.
How does this apply to out tasking and managed and cloud services ? I suggest that early, pre chasm, adopters buy technology on feature / function - these are the kind of guys that say - lets put up a wireless LAN environment this weekend and "Have a go" - but as the technology matures the issue of manageability and reliability becomes an increasingly prevalent buying criteria - so much so that I see Main street buyers putting "Manageability" as a more important buying criteria than Feature / Function. It doesn't matter how sexy the technology is - if I can't manage it then I don't want.
And herein lies the reason why technology vendors need to enable the managed service provider - having a slew of MSPs out there supporting your technology is a very scalable means to overcome the mature buyers concerns and a way to lubricate and speed the way to Main street buyers ... and mainstreet is where the big bucks are.
As I was looking at this lifecycle I was thinking how a enterprise or SMBs purchase criteria would change across that lifecycle as the product or services moved from appealing to Innovators and Early Adopters across the chasm to the Early and Late Majority (also known as Main street) and eventually the Laggards. Each type of organization will have different criteria when selecting technologies and therefore vendors need to create different value propositions depending on where their product or service is in the lifecycle.
How does this apply to out tasking and managed and cloud services ? I suggest that early, pre chasm, adopters buy technology on feature / function - these are the kind of guys that say - lets put up a wireless LAN environment this weekend and "Have a go" - but as the technology matures the issue of manageability and reliability becomes an increasingly prevalent buying criteria - so much so that I see Main street buyers putting "Manageability" as a more important buying criteria than Feature / Function. It doesn't matter how sexy the technology is - if I can't manage it then I don't want.
And herein lies the reason why technology vendors need to enable the managed service provider - having a slew of MSPs out there supporting your technology is a very scalable means to overcome the mature buyers concerns and a way to lubricate and speed the way to Main street buyers ... and mainstreet is where the big bucks are.
Subscribe to:
Posts (Atom)
