Fixing Supply With Demand
Seize the demand day
Labour shortages and extra absenteeism are being reported all over the world. In Australia with fifty-year lows in unemployment and over 400,000 job vacancies, finding staff will be an issue for the foreseeable future. We hear similar issues in countries like the UK and United States. In many industries this has been compounded by other issues like supply chain problems for retail, catch up demand in healthcare or fluctuations to schedules in transport. These issues combined with Covid related absenteeism has made it hard for customer facing operations of all types to hit their target service levels. Long queues and waits are reported in many industries and have had sporadic media coverage. For example, Choice magazine reported average wait times for Qantas of over 39 minutes.
In our recent paper “Twenty ways to handle a crunch” we described some possible supply side mechanisms that may help and using strategies like increased deflection to existing self-service. However, this kind of “crisis” can also be used to provide momentum and incentive to tackle issues that need broader business collaboration in order to drive out contacts or make them easier to handle. Our finding in years of studies is that the overwhelming majority of contacts to manned channels (chat, voice, messaging) indicate some form of problem or confusion. The portion of issues and problems in manned channels has increased as self-service technologies have matured and take up has increased.
In our new book, The Frictionless Organization, we highlight that in most organizations those who
handle customer contact (like call centres or chat teams) don’t cause most of these contacts. In twenty years of studies, we have found that 80-90% of contacts are caused by products, processes or systems owned by other areas of the business. The only contacts that the customer facing areas cause are things like “repeats” where they failed to resolve a problem. However even that is not completely true as it may be issues like business rules or systems access that prevent front line staff solving a problem and those rules and systems are also controlled by other areas. The chart shows a business with very high repeat rates. Deeper investigations showed that it was delays in other areas that drove these repeat contacts as customers called to ask, “where has my X got to”.
Given that tackling this type of demand is a whole-of-business problem, the current “supply crunch” is a one-off opportunity to get the whole of business to focus on demand to relieve the pressure on customer facing staff. Demand reduction strategies produce long term sustainable solutions that save money and make customers happier. Executing these demand strategies will need help from areas that may not see themselves as responsible for customer contact. This current crisis is an opportunity to involve more areas and tackle things previously considered too hard or off limits in a similar way to the amazing pivot that many companies achieved to at home working as the pandemic started in 2020.
For customer facing areas to have credibility when they look for help and support to tackle demand, they need to have a number of things in place. Firstly, they will need credible data on the nature of demand (see our recent paper on “Understand Friction” for more on this). Secondly, they will need senior executive sponsorship to confirm that other areas are accountable and must help. In the book we talk about the measurement and target setting mechanisms that be used to hold each area to account. Thirdly, it can help if the customer facing areas can demonstrate that they are tackling the areas of demand that they control (repeat contacts) so that they are credible when discussing broader demand issue. (We’ll cover that issue of repeat contacts in our next paper).
In this paper we will look at five strategies and related case examples that demonstrate how to tackle the demand side problem quickly and effectively:
Work out who owns what issues and what to prioritize
Fix demand issues driven by products, process and technology
Find some demand answers fixable by supply
Apply all possible digital and automation solutions
Clean up “messy” demand.
Let’s look at each in turn.
1. Working Out Who Owns What and What to Prioritize
The methodology we lay out in the Frictionless Organization is designed to get agreement on who should own each different driver of contact. The first prerequisite is to have quantified data on what is driving contact (which we explained in our paper on Understanding Friction). Then we recommend a workshop process, sponsored by senior leadership (this may even be the CEO) to agree who owns each contact driver and to agree broad strategy for that contact type. One of the “tests” we apply to classification of contacts, is that they do start to make it clear as to which area is responsible. However, we know that it is also a healthy debate and it can lead to the need for further root cause analysis. A workshop process to assign and agree initial ownership of reasons also forces some collaboration and debate and engages all areas with the process. This first cut “assignment” doesn’t have to be perfect and subsequent root cause analyse may identify that another area really owns the issues but it’s a start. In the book we describe the mechanisms to make this kind of ownership stick.
To determine the strategy to apply to each contact type, our suggested workshop process asks whether the contact is valuable to the customer and company or not. This analysis then drives out likely strategies using a 2*2 matrix (see right) depending on how the owners perceive these contacts. For example, those contacts that are irritating to company and customer are candidates for “elimination”. They usually include contacts where things are late, confusing or appear unnecessary. Customers start these interactions with comments like “where is my” or “I don’t understand” or “this doesn’t work”. There may be other contacts that customers value, but the business doesn’t e.g., what’s my balance?” and those lend themselves to automation. The focus of this paper is the contacts that the organization perceives as irritating.
Once the strategy is clear some form of priority setting follows. Priority setting is typically a two-part process. The volume and costs of contacts makes it clear where most benefits can be achieved.
We typically show the “workload” and or costs from largest to smallest. The highest cost contacts are usually perceived as highest priority, but other data can help drive priority setting such as the satisfaction levels associated with a reason, the likelihood of escalation to complaints or the correlation of a contact type with loss of customers. Priorities may also change once the organization explores the feasibility of solutions. The highest cost problem may be the toughest to crack so often priorities are revisited once the business has had a chance to dig deeper on root causes and analyse the cost of solutions. Getting all the areas of the business to take ownership of these contacts, agree a strategy, set priorities and explore solutions can be a very effective way to mobilise a business around customer issues. The next section describes example solutions that have come from this process.
2. Demand solutions through product, process and technology
The solutions to demand problems can take many forms and can be strategic in nature causing a rethink about products and processes. Solutions can also question risk settings and require new or amended technology. In many instances there may be partial solutions that can be done quickly while longer term solutions are put in place. We’ll look at an example of product, process and technology solutions.
2.1 Example Product Solution:
One major telco looked at the data on the rate of contact across its various mobile products and realised that two mobile brands were producing a disproportionate and unprofitable volume of contacts. They were the lower cost products but drove the highest rate of contact. In a ruthless strategy the product heads agreed to drop these brands because the rate of contact demonstrated that these products frustrated customers and lost money. So, the analysis of demand led to a major change in the product mix which in turn dropped the rate of contact overall and reduced customer churn. In a health insurance business, the company called a halt to certain product offers and associated marketing when they realised how many contacts the process was driving. They shifted from offering the first month free to offering the third month free because this allowed payment processes to stabilise, and the contact volume fell away for issues associated with that up front discount. These are good examples of product changes being used to manage “failure demand”.
2.2 Example Process Solution
In another health insurance business, customers who wished to suspend their policies because they were travelling for long periods, had to call in to provide dates of travel and then send some form of proof such as a travel itinerary.
This led to additional contacts as customers wanted to check documents had been received and involved filing and storage of paperwork. This process seemed heavy handed because once a policy was suspended customers couldn’t claim against them. There was almost no risk in suspending a policy and the need for documentation appeared to be a legacy process (see the cartoon). The solution was in two parts. First the company took away the need for customers to send proof of travel. The documents were never accessed or validated so this saved work in the back office and reduced the “did you get my” type contacts. Once there was no need for documentation, it became easier to automate the transaction and the web site and IVR were updated to allow suspensions. Customers just needed to confirm the dates of the suspension and pick a reason. Making the self-service available dropped the demand by 90% for that contact reason and made it easier for customers.
2.3 Example Technical Solution
Discussions around demand can be quite enlightening for technology areas who may be unaware of the impact of some system designs. For example, at one company the most common contact was customers saying they couldn’t log onto the digital solution. This problem was “owned” by the head of digital but the “real root cause” was the complex rules that the legal and compliance team had insisted on for passwords. Customer passwords had to be a minimum length and had all kinds of formatting rules and restrictions. These weren’t even well explained on the web site or to the front-line contact teams. The organization was able to apply a “fix” in two stages. A quick win was to make the explanation of the rules and error messages around these rules more accessible and easier to understand. Part two was changing the rules and that needed a more complex set of analysis and IT changes and discussions with the risk team on the security implications.
3. Demand Answers Fixable by Supply Side Solutions
Staff shortages pervade across businesses but problems in one area often cause issues elsewhere. If key processes like claims or mortgage applications get “behind”, then areas like call centres get besieged with what we call “where is my?” contacts. In this omni channel world, problems in one channel also leak into others so delays in email response lead to more chat and calls. A potential answer to extra demand caused by backlogs and delays in other teams is to temporarily apply resources to fix the backlogs.
In one example of a “backlog blitz” we moved staff from a contact centre to help a team that was in backlog. In the short term this caused slightly longer waits in the contact centre but within days, fixing the backlog in this other team, drove down contact demand and enabled the contact centre to get back to acceptable service levels. We’ve seen similar examples in insurance businesses where backlogs or delays in claims causes a major spike in “where is my claim” enquiries. Re-deploying staff to help fix the claims backlog didn’t mean that they all needed to be claims experts. The “temporary” claims staff were given simpler processes like acknowledging receipt of emails, removing duplicate requests and validating information. That freed up the claims team to apply their expertise but gave the team as a whole, extra capacity.
4. Apply all possible digital and automation solutions
In our paper on “Twenty ways to handle a crunch” we describe promoting and exploiting digital solutions that already exist. However, the supply shortage may be the trigger for new automation investments that were previously off limits. Chat bot and speech bots are being used increasingly to try and automate the simpler transactions and enquiries hitting those teams. The analysis of contact drivers in call or chat teams will make it clear where these opportunities exist. They’ll show up as simpler information or transaction requests such as “when’s my bill due” or “did my payment go through”.
In many examples, chat bots have been able to automate 15-30% of the work that chat teams are handling. There is an upfront cost of development and refinement of these solutions but there is also a hidden customer experience cost. To get this level of automation, usually means that the bot has to be the first point of contact for all these enquiries. So if the bot can answer 20% of the questions this also means that it fails to answer the other 80%. That is an interesting customer experience trade off but one that many organizations are willing to make for the 20% workload saving. Part of the trick is making sure the bots are transparent in setting customer expectations and enable a customer to exit quickly to the manned channel with little wasted effort.
Part automation solutions may have less downside. Robotic process automation (RPA) can be used to automate time consuming processes that free up time for the front-line customer team.
Possible uses include loading data, processing incoming forms, acknowledge receipt of emails or any repeated activity with known inputs and outputs. “Augmenting” the front line in this way provides a permanent solution and time saving.
These types of technology can also be used to “re-set” customer expectations when backlogs and delays develop. Rather than waiting for customers to “chase” items that are delayed, contacts can be reduced by updating customers on delays and expected turnarounds. Of course, some customers will make contact at this point, but many will be grateful for being “kept in the loop”.
Many organizations still have an opportunity to part automate the front of contacts using Call Telephony Integration (CTI) and Call Line Identification (CLI). These technologies “look up” the customer and load systems ready for a conversation. Both technologies help streamline identification processes, but we see many instances where the identification processes haven’t been adapted to exploit these technologies. Other time savings technologies that are maturing fast include things like “co-browsing” where information can be quickly shared with a customer or use of formatted texts and emails to short cut conversations. Sending terms, conditions and rates is a better experience than reading them as the cartoon illustrates.
5. Clean up messy demand
In many organizations there are other forms of demand that are unintended or messy. Customer enquiries often end up in the wrong team when there are many options to choose from (e.g. complex skill models) and in some companies a large proportion of queries fall outside the scope of the organization. In one government department we sampled recently, 60% of callers were for other departments. Messy demand can also happen in branch and service centres where customers go to the wrong counter or don’t have everything needed. In the worst example we’ve seen 45% of contacts were transferred between skills because the customer landed in the wrong place by accident. In another business we saw misrouted back-office work account for over 20% of work items. We’re not saying all transfers and hand offs are bad, but they usually are when they occur more by accident than design.
One temporary mechanism that can increase capacity is a “human switchboard” or a “concierge” in retail. Service NSW uses a concierge in their service centres to direct customers to self-service, to validate they have the information needed and to direct them to appropriate counters. This one person deflects some customers to self-service and stops customers queuing where they don’t have the information to complete the process. It’s a great example of using a person to “clean up and divert” demand. Some companies (most famously AAMI) have always favoured human switchboards over IVRs and speech-based menus.
This can also be a way to mobilise a group of temps or students with limited training to help in a resource crunch. They are set up with the questions needed to route work more accurately, to deflect work to self-service or to “triage out” contacts “outside scope”. The processes are simple and can be trained in days. These models can also be extended to defer work to other parts of the day. We’ve had models like this up and running in two weeks and in one instance they reduced demand by 20% and took the pressure off the front-line teams through a mix of triage and more accurate routing. They can also be used to create temporary capacity while more complex demand problems are addressed.
We’ve described some “band aid” temporary solutions to “messy demand”, but should also note, that if customers consistently land with the wrong staff, it may be time to revisit both the navigation mechanisms, e.g. IVRs or menus, and the skill model that is being used. We have seen many organizations where skill models have become so complex that customers will struggle to navigate them and they are very complex to manage. One organisation had 65 skills which in effect meant they had 65 rosters to manage! We have written other papers on alternative skill models such as “tiered models” where customers land in a group that handle simple enquiries and are quickly triaged to specialists or deeper tiers. These models have the potential to remove “messy demand” but only if processes are also re-aligned, which should be the subject of another paper.
With labour shortages likely for the foreseeable future, now may be a great opportunity to tackle the demand for certain services. It can also be a great chance to mobilise the whole business around these problems and create solutions that now appear more attractive. It is always surprising to us how investing 4-6 weeks and a few SMEs inside the business on a demand design sprint can crack some big issues quickly to save demand, reduce cost and improve the customer experience. The techniques we have described are often more complex than we can describe in a short white paper, but we are always happy to discuss them in more detail. For more information email us at email@example.com or call 03 9499 3550 or 0438 652 396