Even though folks may be familiar with differences between B2B and B2C, here’s a quick refresher.
B2C companies are Businesses that primarily cater to end-Consumers (Examples: Amazon, Walmart, eBay, Uber, Pepsi etc.). B2B companies are Businesses that primarily cater to other Businesses (Examples: Cisco, GE, Salesforce, etc.)
Customer Support for both is markedly different —
In summary, B2C companies usually deal with large volume of cases with low complexity while B2B companies deal with lower volume of cases with high complexity.
What is perhaps the only similarity between the two is the business need to store/record the issue (case)filed by the customer, and the corresponding workflow to resolve it. This requirement is met by case management systems like Zendesk, Salesforce Service Cloud etc., These systems are also known as Customer Support/Service CRMs (Customer Relationship Management; though CRM terminology has been abused¹ here by the vendors) .
These systems are also the “System of Records” for Customer Support (here’s an excellent post that explains the difference between system of record vs system of engagement vs system of intelligence. It is a must-read if you are not familiar with this terminology).
These workflow-based systems are actually a huge business! Salesforce alone generates $2.8B annually in revenue by selling its Support Cloud. In 2016 that number was $1.8B.
Zendesk made $430M with an impressive revenue growth!
… just on the back of a ticketing system. Jeez!
One may (rightfully) be compelled to draw a parallel to ITSM (IT Service Management — these systems store the cases raised by internal employees for IT to resolve) or even Bug Tracking systems (like JIRA from Atlassian used by engineers to track software bugs). ITSM, Bug tracking and Support CRM systems are essentially ticketing systems but built for different users.
It is instructive to note that ServiceNow (leader in ITSM) generated $1.9B revenue in 2017 and has recently entered the world of Support CRM. ServiceNow had an interesting² journey since its inception.
Let’s come back to the differences between B2C and B2B customer support.
Since the nature of issues in B2C is relatively simple, it is a perfect candidate for automation. Combined with the fact that machines can better cost-effectively handle large volume of cases versus humans makes the case for automation very strong (and achievable) which usually presents itself in the form of self-service.
Examples of this are:
- Chatbots. e.g., Erica from Bank of America,
- RPA (Robotic Process Automation) —allows automating the entire process of logging to different systems and grabbing right information at literally click of a button. Best way to see it is in action… youtube video from IBM.
- Auto-suggest answers from FAQ as you start typing your question.
Automation in B2B customer support is, however, trickier as the issues are more complex. Consequently, B2B support has traditionally not been considered a good candidate for automation.
Moreover, since the customer was required to pay handsomely for support at the time of purchase (unlike B2C model), sellers could justify hiring humans to help resolve the cases and it was good money as well (For hardware-centric enterprise, typical revenue split across product/service is 70% / 30 % with services usually bringing in higher margins).
Intuitively then, B2B support should be more personalized/kick-ass as compared to “automated” B2C support, but in reality it is quite the opposite. B2B companies are less customer-centric than their B2C counterparts. McKinsey quantifies this with customer-experience index ratings:
B2C companies typically score in the 65 to 85 percent range, while B2B companies average less than 50 percent.
Why is that the case?
- Complex Software — In agile delivery world, software is constantly being updated and also becoming more complex making it increasingly difficult for the support staff to keep up with the pace of changes while simultaneously supporting existing software. Result is increase in MTTR (Mean Time To Resolution) for resolving cases which results in poor customer experience.
- Support Staff Churn — Support is mostly a thankless job and agents consistently report a lack of purpose which translates into high turnover (TSIA reports 15% churn and increasing). Experienced support agents also express job dissatisfaction in being required to handle mundane support cases which ideally should be handled by machines (cheaper, faster, scalable). In my research — more on that later — I didn’t find any broadly deployed intelligent B2B self-service solutions; they would be a win-win-win for vendor (cheap)/customer (faster service)/agent (frees up time to work on interesting cases). Combined with the fact that it takes roughly 1–2 months to completely train a new agent, the problem of not having well-trained and motivated agents is compounded. All this flows downstream into a less than satisfactory customer experience.
- Support cost — Because of the complexity of B2B cases, it costs the company anywhere from $250 to $1000 to service each case on average. Companies initially start with support staff as full-time employees but to reduce support costs, they later decide to split it in different layers — Level 1 support is the first line of support, which is usually contracted out. Their job is to resolve simple cases and do “case routing”— decide which sub-team in Level 2 support is the right owner for the case. The cost per case at this level is around $250 per case. Ironically, this is also where customer experience takes the worst hit — Level 1 agents usually have little knowledge about product let alone the customer. This creates a very undesirable side affect — CEOs don’t realize that even though cost to service a ticket overall decreases in this model, the decrease in customer experience will have much bigger detrimental affect later, in terms of increasing customer churn and loss in renewal revenue.
- Low-priced B2B products —In the case of low-priced (sub $10K) hardware products, high cost of support tends to wipe out the entire margin on just 1 or 2 support calls. At Cisco, I led a product portfolio of firewalls where our lowest end firewall was priced at around $5K. They were usually bought by SMBs which usually didn’t have a well-trained tech team to manage the devices — this meant, everything from documentation to device configuration and even product packaging needed to be simple to avoid any support calls. However, such products are sold to a larger number of customers (vs high-priced products being sold to fewer enterprises) and in the absence of any technology solution to auto-resolve such cases, companies have strict requirements for such cases to be only serviced by Level 1 support resulting in poor customer experience.
In a hardware-centric model, customers paid upfront both for hardware and support (as a one or multi-year contract) i.e., the vendor ended up receiving all the monies upfront and it also created a high switching cost for the buyer. In other words, there was no incentive / burning desire on part of the seller to provide the best customer support. Instead, it was treated as a necessary evil, cost of doing business. Forget customer experience.
However, with the transition from hardware to SaaS, CEOs can no longer ignore customer experience. (How SaaS transition is impacting a different business process — Quote-to-Cash — is covered here).
First, sellers don’t receive entire money upfront (either for product or for support); essentially the revenue model has been turned upside down. Sellers are not only required to maintain similar staffing level for support but also to provide superb and even differentiated customer experience. Combined with the fact that it typically takes an year for the seller to realize cost of that sale, if that customer files any support cases in the first year, seller is already in negative margin territory. That’s a double whammy!
Second, the switching cost for the customer has now become low (due to SaaS nature of the product). If the customer is not happy with the service, they can now easily switch to another vendor who is providing better service. In other words, initial sales is and the vendor needs to provide continuous value otherwise the customer will churn³.
This has forced CEOs to now think of customer support as not just an ancillary function, but an essential function needed for fueling company’s growth (similar to sales). Unless the seller is providing the best post-sale customer experience (which is usually in terms of customer support), the customer is likely to churn.
In other words, customer experience is back baby.
Google Trends agrees as well:
Furthermore, with the consumerization of enterprise, B2B customers now expect faster resolution and even proactive / predictive / preventive support.
Suffice to say that now,
Support is the new Product
How are sellers reacting to the new challenges in Customer Support?
There are two aspects to post-purchase customer experience:
A new business function — incorrectly (more on this later) called “Customer Success” — has emerged to solve for on-boarding challenges.
In fact, Google Trends shows a strong correlation between “customer experience” and “customer success”.
The rationale is following — Customers are more likely to churn in the first few months of the purchase. This happens because enterprise products are complicated and it requires training and expertise to be able to get the maximum value from these systems. In pre-SaaS world, the onus for getting that training was squarely on the customer (due to high switching cost of the hardware product). In SaaS world however, the on-boarding burden is now seller’s to bear and that has given the rise to creating a new business function called “Customer Success”. People performing this function are called CSMs (Customer Success Managers) and the tool they use for doing their job is called, well, “Customer Success Platforms” (google search).
But, here’s the rub.
One would expect a customer success solution to do exactly what it claims i.e, help customers to be successful i.e., help them them grow their business (if your product/service is merely filling a functionality gap, you are likely to be switched out eventually by a vendor who has a laser focus on their customer success than their own). However, customer success is currently boiled down to just customer on-boarding⁴ — which is beginning of the journey, but certainly not the end of it. In other words,
True Customer Success Solutions don’t exist today.
Resolving ongoing issues
This is where the bulk of trouble arises as existing tools (ticketing systems from earlier in the article) have just been limited to being system of records and not being system of intelligence. They don’t provide ability to either stop the cases from being filed (case deflection) or allow agents to solve the cases faster (reduce MTTR). Agents are forced to play Sherlock with every new ticket being filed; they have to log to multiple systems (knowledge base, emails, dropbox, google drive, bug tracking system and in some cases even code repositories) to find relevant information to resolve a ticket.
I wanted to understand what support leaders/agents think about it, so I went to LinkedIn and started my market research. I spoke to roughly 50 people in B2B support organizations (technology sector; both hardware & SaaS) and they echoed lack of the right tools to improve customer experience:
- “I cannot grow my support team in proportion to my customer growth. I need to make my organization efficient but there don’t seem to be right tools in the market for that”
- “Our Level 1 support is working on a lot of repetitive tasks. We may have 2 people sitting right to each other and working on similar cases.”
- “We routinely lose money on support on our low-end products.”
- “I need to be able to see customer configuration, history of purchase, prior cases in one place.”
- “I would like to make our customer support proactive instead of reactive”
- “We have a lot of data on our customers and we want to apply better learning on that data to serve them better.”
- “Training customers doesn’t help. Our Tier-1 customers usually outsource operations to third parties who are trigger happy in filing basic support questions. How do I automate that?”
- “Since transition to SaaS, I have lost support revenue. I want to offer virtual resident engineer as a value-added (revenue) services.”
- “ I don’t feel excited about my job.”
What are existing vendors (of ticketing systems) doing about it?
Mostly the innovation seems to be centered around support chatbots. I didn’t come across any successful B2B chatbot examples though. Also, there is general industry fatigue around them due to over-promise and under-delivery. Gartner says:
By 2020, 40% of bot/virtual assistant applications launched in 2018 will have been abandoned.
In terms of specific products, there is IBM Watson and Salesforce Einstein but neither of them came up as viable solutions in my conversations. From offline research, it seems that they are also more suited towards B2C than B2B. ServiceNow has done some related acquisitions in this space (Parlo, May 2018; FriendlyData, Oct 2018) but once again impact on B2B support is unclear.
What’s the fix?
To bring B2B Customer Support into the next-generation, here’s my take on changes that it would take both on technology and people front.
System of Intelligence
First and foremost, one needs to start by building a System of Intelligence that has both customer and vendor context. Customer context — who the customer is, what have they bought, what features are they using, logs corresponding to their use of product etc. Vendor context — internal sources like knowledge bases, prior cases, bugs, CRM, ERP, etc. to create an Enterprise Knowledge Graph.
Based on my survey, all support organizations reported that roughly 20%–30% of incoming cases are transactional cases (e.g., what are my license keys? or how do I do x? ) while the rest (70%–80%) are non-transactional — they require involved effort from humans.
The system of intelligence can help in both cases.
Build a proactive support system (True customer success system)
Existing support model is break-fix or reactive i.e., customers contact the vendor after things break. However, with SaaS, the vendor should technically be able to know when the problem is about to happen (predictive support) or has already happened (proactive support), inform the customers proactively, and then fix the issue. In this workflow, customer didn’t even have to initiate the support workflow.
To enable this kind of support, a system needs to be built that is constantly monitoring what each customer is doing (while maintaining privacy), analyzing any anomalies by cross-referencing data across other customers and then surfacing issues. This is not an easy task but can be done.
Re-think ticketing systems
Existing systems are designed for use by support agents only. They have completely missed the perspective of other business functions i.e.,
- Sales — What is the health of my customer post-sale?
- Product Management — What are the issues are customers keep running into?
- Engineering — Collaborate effectively with support to resolve the case.
- Documentation — What features are most confusing?
- Marketing — Understand customer use cases, their voices. Tapping the really happy ones for case studies.
- Executive — Which customers do I need to spend my time on?
- HR — ok, just joking now ;)
Currently, there is a trifecta of active customer-facing teams — Sales, Customer Success and Customer Support. In some companies, Customer Success comes under Sales organization, while in others it sits within Customer Support organization.
Since entire business case for SaaS rests upon renewal, let’s look at that more carefully.
Sales teams work actively, well, during pre-sales. Customer Success teams (CSMs) get in action only post-sales and help in on-boarding (Issue #1: CSMs have no idea on what was sold/promised to the customer). CSM is largely a non-technical role (akin to post-sales Account Manager). After on-boarding Customer Support teams (highly technical) take over conversation completely and work with the customer during the year on any issue. (Issue #2: There is no formal handover of accounts between CSMs and Customer Support teams).
Come renewal time few months down the line, if the customer is happy, all is good. If however, customer is unhappy, it is likely that finger-pointing will start between the three teams:
Sales Engineer — “What did you folks do to make the customer so pissed?”
CSM — “We did on-boarding well (check the emails) but looks like we didn’t meet all the requirements of what was promised to them.”
Support — “Customer was too demanding and was using the product in non-recommended ways. Can we ensure that we follow the best-practices around configuration?”
Issue #3: Who has the ball?
Issue #4: While Customer Success and Customer Support teams actions have a strong and direct impact on customer’s renewal decision, they are not comped on it. Instead, sales teams will get comped on renewal even though their involvement during post-sales was minimal.
Issue #5: Success metrics for both CSMs and Customer Support are misaligned. For Customer Success, metrics is around successful on-boarding (loosely defined) and for Customer Support it is # of cases resolved/day. Companies do measure NPS (Net Promoter Score) score for each customer independently but that is not strongly tied to either functions.
Here are some recommendations on a possible way to fix this:
- Customer Success and Customer Support Teams should be merged and sit within overall “Customer Experience” organization (don’t call it Customer Support). Names do have a significance.
- The structure of this organization should mimic that of Sales organization i.e., CSM is post-sales Account Manager while Support Agent is post-sales Sales Engineer. In hunter/farmer analogy, Sales organization will primarily be the Hunter, while Customer experiences will be the Farmer.
- Renewal comp structure should be changed. 70% or more commission for renewals should go to Customer Experience teams while remaining should go to Sales (Sales is overall responsible of owning the account and also land-then-expand opportunities). Sales comp structure for SaaS businesses is a complex topic; however, during my survey I found that most companies continue to comp the sales the same way as they would in a hardware (one time sale) model. This was true for both startups and public enterprises (surprisingly).
- Formal handover of account from Sales to Customer Experience teams.
In summary, start treating your support organization as a semi-sales organization. I am not aware of any successful/failed examples here. Please comment or write to me if you know of any such cases.
Gartner indicates that customer service and tools market is worth an estimated $6B. This is just for case management! If you account for the salaries of support staff, which is the total addressable market for an intelligence-based solution, we are talking about a $70B+ market as conservative⁵ estimate. In other words, intelligence-based tools market is worth 10x or even 20x more than the ticketing tools.
However, downside of selling intelligence-based tools is that they don’t scale as easily as cookie-cutter products like ticketing systems, and most important, intelligence is only as good as the data it is being fed.
But once they work, they are virtually irreplaceable.
They are the Tortoise.
- CRM stands for Customer Relationship Management which would intuitively mean a system that manages the entire relationship a business has with its customer e.g., what has customer bought, what trainings have they done, communication history, what issues have they faced, etc. Think of it like a true customer 360 platform (many vendors abuse this terminology as well, in that they provide only a partial customer view, say customer 45, but claim to provide full customer 360). However, current usage of CRM terminology represents only transactional workflow systems (not relationship) e.g., Sales CRM systems record sales transaction , Marketing CRM record lead qualification transaction, and Support CRM systems record customer issues.
- If you are not familiar with ServiceNow’s journey to success, their story is quite fascinating — Fred started the company when he was 50! Forbes has an excellent coverage here. One of the reasons I find their story unique is that it is one of the few cases where technical founder side-stepped from running the company (as a CEO) to have seasoned business folks run the company (not only from a role perspective but, more important, giving them the freedom to act independently). Running a scrappy 5 or 50 people startup is very different than running a 1000 people company and it requires completely different mindset. Some founders graduate to that level, while other’s don’t; they have hard time letting go and eventually their company gets stuck in an orbit never finding that escape velocity to get out and grow as a business. Successful examples of the transition — Microsoft, Google. You can guess the failures.
- Most CEOs of early stage startup companies keep investing heaving into revenue business function (sales) while ignoring customer support. That’s understandable — VCs look for upward revenue growth (especially if you are fixated on 40% rule) and “immediate” growth comes from getting new customers. The affect of not focusing on customer support, however, will come up soon enough during the renewal phase. Unhappy customers leave or “churn”. Controlling churn is the most fundamental, acknowledged and yet often ignored metric for SaaS businesses. Internet is full of literature on this subject but the best one I found was from SaaS Capital (PDF, downloadable upon submitting your info). It explains clearly correlation between churn and valuation and summarizes — “for every 1 percentage point increase in revenue retention, a SaaS company’s value increases by 12% after five years.”
- Most customer success solutions identify new accounts that have not successfully on-boarded — they do that by collecting and correlating information from Sales Cloud (who are the new customers), Service Cloud (have they filed a support ticket) and Product Login (how frequently are they logging in to the product) and generating lists of “at-risk” accounts. CSMs take these lists and an auto-generated email is sent — “Dear customer, Thank you for your recent purchase. If you have questions on how to use the product, here’s a link to some videos. If you’d like to speak to someone, click here to schedule a call”. The call is usually handled by a Sales Engineer. CSMs are essentially acting as post-sales Account Managers. Even though this approach certainly helps in better on-boarding, it is in fact a band-aid approach and is likely to fail in long term.
- I did a LinkedIn Search for “customer support” with following filters:
Industries: Computer Software, Internet, Telecom, Information Technology & Services
Connections: 1, 2 and 3+
Assuming, median salary of 150K, this is a $73B market (490K * 150K). Accounting for other verticals (financial services, healthcare etc.), this number would be easily $100B+.