Preparing your customer service operation for a stress-free retail peak

With the peak trading period rapidly approaching, it’s imperative that retailers set their customer service operation up for success. Not only to deliver fantastic customer experiences that set you apart from the competition, but also to provide a stress-free period for those at the coal face of delivering the service.

We’ve pulled together our top tips for how customer service teams operating in the retail space can best prepare for the upcoming onslaught...

1. Make it easy for customers to contact you

We’re big proponents of self-service and wholeheartedly believe that conversational AI can enable customers to get the answers they need to common and simple questions. This also has the benefit of freeing up agents to work on issues that require the human touch.

But (and it is a BIG but), this must be done at the right time, and with context of why the customer is contacting you in mind, at all times.

When a customer really wants and needs to contact you, it is your responsibility to make it as easy and frictionless as possible. Examples of where customers might need to speak to you include more complex issues such as missing parcels/items, faulty products, and return/refund delays.

a) Publish your contact channels visibly on your website

I’m always amazed by how many retail websites hide away or make you run a proverbial gauntlet just to find the customer service telephone number. All this does is create frustration.

Make all your contact channels easy to find on your website.

b) Simplify IVRs and offer direct routes to human agents

Now that your customers have easy access to your contact channels, don’t put them through IVR hell in order to speak to an agent. Whilst it’s necessary to have some level of IVR to best direct the customer to an agent that can help, keep it simple with no more than 2 or 3 levels at the most.

Better yet, if you’re able to predict why the customer is calling and route to an agent, do that! The same applies to webchat if you offer it: Don’t make the conversation unnecessarily convoluted in order to speak to an agent.

2. Manage queueing expectations and enable callbacks

In spite of your best planning efforts, you will likely to end up with unexpected peaks in contact, and when this happens it’s imperative that you set your customer expectations, as this will negate a lot of frustration.

a) Inform customers of the accurate wait time

Whilst this is commonplace for most voice-based contact, you should do this across all channels including asynchronous communication, and keep that wait time up to to date. If during the course of peak, it’s now taking you 48 hours to respond to email rather than 24 hours, make sure this is reflected in any messaging on your website or autoresponders.

b) Offer callbacks for customers on the phone

If you offer voice as a channel, offer the customer a callback whilst in the queue, or better still, allow them to pick a date/time that best suits them.

3. Enable self-serve to reduce wait time and deflect low-value contact

Sometimes customers just need reassurance that their parcel is on it’s way. Other times it’s just impatience (we’ve all been there). However, a customer contacting you and getting through to an agent just to double check when their parcel will be delivered isn’t a good use of anyones time. Particularly when it can be easily served by implementing AI and Automation and with platform accelerators to get you started.

4. Enable agent ownership of a customer problem

In a recent study by Gartner, one of the biggest complaints customers had was needing to repeat themselves. This is typically caused by being passed between multiple agents on the phone or chat (”I’ll just transfer you to a colleague who can help”), or for email and social channels where multiple messages are required, and the response ends up with a completely different agent.

Not only is this all very frustrating for the customer, but there is a massive inefficiency cost to you as an organisation, as each “transfer” requires the next agent to get up to speed on the issue and cover the same ground that another agent already has.

Customers love it when a person takes ownership of their problem - it builds confidence that someone cares, which matters more than you think, and also creates a more personal relationship between the customer and you as a retailer. All of this contributes to future loyalty and the customer purchasing from you again, or recommending you to friends and family.

a) Track, monitor and reward based on First Contact Resolution (FCR)

A lot of customer service operations focus on Average Handle Time (AHT) as one of their main KPIs. The thought process being, the shorter the engagement, the more engagements an agent can handle per hour. Makes sense. However, this can drive poor agent behaviour when incentivised only on this metric as the agent will be more focused on getting the engagement completed as quickly as possible, rather than solving the customer's issue.

Also, if you are employing AI and automation to handle high volume/low value contact, you will have likely cut out the “easy” engagements, and those that are left are more complex, and by their nature take longer. We’ve seen with our own customers, that AHT actually increased in some cases for human agents once AI and automation was implemented as they didn’t have the “easy” engagements lowering their average.

b) Route responses back to the same agent

Where customer issues require multiple engagements, route those engagements back to the same agent. This applies to all channels, but particularly asynchronous messaging such as email and social where the conversation can be more stop/start. Ideally, this would also be employed cross-channel as well such as where a customer starts via email but then phones in to follow up.

An additional benefit of this re-routing can be a behavioural change for some agents who may typically try to avoid more complex engagements, and hope that someone else picks up the response.

Once the agent knows that customer's response will keep coming back to them, they will be more inclined to dig in and solve the issue (Sticks and Carrots).

c) Offer a call when the engagement was initiated on another channel

If an agent is engaging with a customer on a non-voice channel, encourage them to offer a call with the customer. This will cut through the back-and-forth messaging driven by certain channels, will likely resolve the customer's issue more quickly, and will really stand out as a positive experience.

5. Review and optimise your self-service, automation and knowledge bases

All retailers will be making use of FAQs and many will have some form of automation, be this a chatbot/voicebot or automated email responses. You may also have resources specifically for your agents such as canned/template responses or knowledge base. It’s imperative that these are all reviewed and optimised in advance of and during peak as key information such as delivery times and returns windows often change during this period (e.g. you often have until the end of January to return items ordered in December).

There’s nothing more infuriating then getting conflicting information as a customer, and as a retailer there’s no point going to the trouble of providing self-serve and automation options if they aren’t correct. You’ve missed the opportunity to save your agents valuable time.

6. Proactively communicate when there is an issue

Things will and do go wrong, it’s a fact of life. Be that an issue at the distribution centre packing the wrong item, courier delays, or packages lost altogether. You can’t control that.

However, what you can do is control how you communicate an issue to a customer. And customers prefer you let them know there’s a problem before they find out by having to contact you. So make sure you’re on the front foot and let them know their order is delayed and importantly when it will be delivered. It will save you a world of pain later on.

7. Plan the right agent resource for your expected peaks

Getting resourcing right during peak is like walking a tightrope. Too little resource and you’ll have a horrible peak of stress, customer complaints, missed KPIs. Too much resource and you’ve lost precious margin that you could be investing elsewhere in your operation.

What you need is “just enough” to handle your planned volumes, plus a little slack for those unplanned/unforeseen issues.

But how much resource do you need...?

a) Leverage Contacts Per Orders in your forecasting

Historic contact volume data can be a good indicator of what might happen again this year, but purely looking at contact volumes might not provide enough granularity. Instead, you should broaden the data you analyse and determine how your Contacts Per Order (CPE) which is the Total Contact Volume divided by the number of Orders you received in a given period.

CPE = Total Contact Volume/Orders

You can look at CPE broken down by peak vs. non-peak, by channel, weekday vs. weekend, etc.

Once you have determined your CPE figures in the granularity you want, you need to speak with your Trading Team whom should have forecast the expected number of orders you a likely to receive during this period. Then it’s simply a case of multiplying the CPE by the Forecasted Orders to calculate you Forecasted Contact Volume.

Forecasted Contact Volume = CPE * Forecasted Orders

For example, if you have a CPE of 2.3 and Forecasted Orders of 1,000,000. Your Forecasted Contact Volume would be 2,300,000.

b) Use a dedicated Workforce Management tool

Whilst we know a number of operations successfully run resource scheduling using spreadsheets, this typically involves significant human effort, and can create a little panic and overhead when those unforeseen issues occur, and schedules need to be changed quickly on the fly. For these reasons, we recommend organisations with 100 or more agents to use a specific Workforce Management tool.

Amazon Connect users should certainly look at Amazon Connect Optimization which provides native forecasting, capacity planning and scheduling capabilities all within the same interface and on their consumption based pricing so you never pay for more than you need.

Whilst Zoom Contact Center customers need look no further than the recently launched Zoom Workforce Management. There are also a number of contact centre platform-agnostic providers out there such as Calabrio.

8. Understand why customers are contacting

Let’s be honest, the customer service team is rarely the cause of the contact it receives. It’s normally driven by problems elsewhere such as the website, distribution centre or courier.

You’ll certainly know about these problems once they really hit, but could you know about them in advance?

a) Deploy Conversational Analytics to identify and flag problems

Conversational Analytics uses AI and Machine Learning to identify common issues, trends, and areas for improvement within your customer service operation. Platform agnostic and best in class solutions like Xdroid have been leading in this space for many years. AWS has also been maturing its Contact Lens offering with out-of-the-box integration with Amazon Connect.

Using tools such as these enable you to spot problems and adjust your communication strategy, such as proactively communicating delays in parcels or removing problematic products from sale.

9. Scale for peak

Are your contact centre platform and engagement channels scaled for peak? This may sound a strange question in the world of cloud, but limits still exist, particularly where the digital world meets the physical world.

a) Review and scale capacity and licenses

Depending on your contact centre platform and what channels you have configured you’ll likely need to consider scaling up either your user licenses and/or concurrent contact volumes.

b) Monitor and optimise your costs during peak

So now you’ve scaled up for peak, it’s time to forget about it right? Wrong!

There's usually at least some element of consumption-based cost (voice minutes), and if you’re not careful these costs can spiral, leaving you with a nasty surprise bill at the end of peak.

The suggestions above (such as implementing callbacks within queues) will certainly help reduce costs, but it’s important to review and optimise the costs of your operation as much as you can. For those using Amazon Connect where all costs are consumption based, we provide a Cost Optimisation solution that will ensure you’re fine-tuned for peak.

10. Don’t forget your people!

Finally, but most importantly: Don’t forget your people!

All too often we focus on the customer experience and ensuring we at the very least service them, but ideally delight them. However, your single most important asset during peak (and also not during peak) is your people. They are the ones on the front line making sure customers get their responses and will dig in with you when those unforeseen issues occur.

Remember to say thank you and also see if you can bring some fun into what will hopefully be a very busy time for everyone!

Summary

Whilst retail peak can be a challenging time for customer service teams, with the appropriate planning you can not just survive, but actually thrive, offering outstanding customer service that stands out from your competition.

Still worried about peak? Why not watch our on-demand webinar to learn about how you deploy Zoom Virtual Agent before peak to enable self-service and deflect low value contact without the need to re-platform from your current contact centre provider.

Can’t make the webinar or just want to chat to someone about your concerns? Get in touch, we’d love to help find solutions to your challenges.