Service level agreements (SLAs) are policies you define that specify and measure the response and resolution times your support team delivers to your customers. There are many standards of service you can define in an SLA policy, including quality, availability, or timeliness. Of these SLA metrics, the first reply time metric is of great importance to your customer's overall experience and the metric that can be most influenced by omnichannel routing.
You can use Explore reporting to understand your current metrics, identify your first reply time SLA goals, and then use a combination of omnichannel routing features to achieve those goals.
- Gathering and monitoring your first reply time data
- Understanding the first reply time SLA metric
- Creating a first reply time SLA policy
- Defining omnichannel routing queues that align with your SLA policies
- Optimizing your omnichannel routing configuration for your SLA policies
- Other optimizations to support your SLA policies
Gathering and monitoring your first reply time data
Before you implement omnichannel routing or reply time SLAs, it's important to understand your service levels per channel. This data serves as a benchmark against which you can measure the success of your omnichannel routing configuration and SLA policies.
In Explore, you can generate reports on your reply time data for email and messaging tickets. When defining your report, you can specify the time unit (seconds, minutes, or hours) and whether the metric calculation is limited to business hours (if configured for Support or messaging) or uses calendar time instead. Furthermore, you can decide whether you want to report on the average first reply time or the medium reply time, and you can apply time filters to look back at historical data. For more information, see reporting on first reply time for Support and messaging tickets.
To measure a reply time for calls, you should use the call wait time metric. This is the metric that most closely aligns to the first reply time metric used for email and messaging tickets because it measures the time elapsed between then the call is received and initially routed until the customer's first connection to an agent. For more information, see Inbound calls by wait time.
Understanding the first reply time SLA metric
-
Queue timeThe amount of time tickets spend in the queue before being assigned or offered to an agent is directly related to three things:
- Agent availability. This is determined by the groups of agents the queue routes to and the status, capacity, and skills of the agents within that group. When using custom omnichannel routing queues, the primary groups of agents are considered first, and secondary groups only when necessary. The more agents that are eligible to receive work from a queue, the faster tickets will leave the queue and be assigned or offered to them. Consider your staffing and scheduling to ensure there is an adequate number of agents in a queue's groups to handle the expected volume of tickets, and address inadequate staffing situations rapidly to maintain desired service levels.
- Ticket position in a queue. Tickets at the front of the queue are assigned first, and queue ordering is determined first by ticket priority and then by timestamp.
- Queue priority. When agents receive work from more than one queue, the tickets from the queue with a highest priority are assigned to agents first. To minimize conflicts, be mindful of the overlap in the primary and secondary groups you configure for each queue and how they might overlap.
-
Acceptance time
In the standard omnichannel routing configuration, email tickets are automatically assigned to agents so there is no concept of acceptance time. However, messaging tickets are offered to agents rather than directly assigned. If the first agent the ticket is offered to doesn't accept it, omnichannel routing continues to offer the ticket to available agents until someone accepts it.
If you're concerned about how long it's taking for messaging conversations to be accepted, you can turn on auto-accept for messaging in your omnichannel routing configuration. This behavior eliminates the acceptance time.
-
Response time
After a ticket has been assigned or accepted, the agent must read and understand what the end user needs. Then they must write an appropriate response. You can minimize the response time by using the Advanced AI add-on to summarize the request and even respond to tickets.
To learn how your omnichannel routing configuration affects the first reply time metrics, see Optimizing your omnichannel routing configuration for your SLA policies.
Creating a first reply time SLA policy
- Existing plus X%: With this model, you decide the percentage by which you want to improve your existing service level. If you and your customers are quite happy with the existing levels of service, you can still use this method to incrementally improve your standards.
- Industry standard service levels: In some cases, there are clearly established service level standards that you want to achieve. A common example is a standard of 80% of messages being responded to within 20 seconds.
- Differentiated service levels: This model is a good fit if your services levels vary by your type of end user. For example, you allow users to pay for a premium service level or you provide higher priority and service levels to important customer segments.
In most cases, you'll want to define different first reply time policies for each channel your customers use to contact you. Typically, a first response time for email tickets can be counted in minutes or hours, versus live channels where a first response time should be counted in seconds or minutes. Beyond channel-based service levels, you might also differentiate service levels based on subject matters, individual end users or organizations, or even customer sentiments (if using the Advanced AI add-on). Sensitive topics, highly valued customers, customers at risk of ending their business relationship with you. or customers who are angry might warrant higher service levels.
For example, you might define four SLA policies that are defined by two things: the channel by which the ticket is received and whether or not the ticket is from a VIP customer. In this case, your list of SLAs, ordered by shortest reply time to longest, might be: VIP Messages, Non-VIP Messages, VIP Web Form, Non-VIP Web Form.
To create an SLA policy, see Defining SLA policies.
Defining omnichannel routing queues that align with your SLA policies
A key way to maximize the benefit of using omnichannel routing to achieve your SLAs is to align your routing queues with your SLA policies. Omnichannel routing inherently orders tickets by priority and creation timestamp within queues. That means the tickets that are closest to breaching your next reply time SLAs are typically at the front of your queues, and therefore will be assigned first. When the standard queue ordering behavior is combined with queue priority and ordering that matches the SLA policies, you ensure that the queues containing the most important tickets that require the shortest response time are prioritized at both the queue level and within the queues themselves.
- Have the same conditions as your SLA policies
- Are in the same order as your SLA policies
- Have the appropriate priorities that match the stringency of each SLA policy
Continuing with the previous example SLAs, you might create four custom omnichannel routing queues of the same names and conditions and order them in the same way. In this scenario, the VIP Messaging SLA requires the shortest reply time, so the VIP Messaging queue is given the highest priority and ordered at the top of the list.
In these queues, there are many ways you might choose to configure the groups the tickets are assigned to. However, for the sake of this example, let's say all four queues route to a primary group of Tier 1 and, when necessary, a secondary group of Tier 2. Another option would be to only specify secondary groups for the VIP Messages and VIP Web Form queues.
Optimizing your omnichannel routing configuration for your SLA policies
When optimizing omnichannel routing settings for your SLA policies, there are two primary considerations outside of queues: agent capacity and the individual routing configuration settings you use.
Aligning agent capacity with SLAs
When using omnichannel routing, the capacity rules established for agents directly affect their availability to receive, and thus reply, to tickets. The maximum capacities determine how many concurrent tickets an agent can be assigned per channel. When agents have capacity rules, omnichannel routing keeps track of every open, active ticket assigned to each agent and only routes work to agents when they have an eligible status and are below their maximum capacity for the channel.
Capacity rules for calls are easy because you only have two options: zero, meaning the agent can't receive calls at all, or one. Getting capacity rules right for email and messaging channels can be trickier. Setting the maximums too low can have a detrimental effect on service levels because tickets get stuck waiting in the queues too long before an agent can reply to them. On the other end of the spectrum, setting the capacity too high can also be detrimental because agents could be so overwhelmed with the work assigned to them that they can't get to all of their tickets in a timely manner.
As a starting point, you can use your intuition to set maximum capacities. However, the best way to determine your capacity rules is to use Explore reporting and take a data-driven approach. When using Explore to inform your capacity rules, look at the volume of tickets created over a period of days or weeks and the number of tickets solved by an individual or group of agents over that same period. If necessary, you can get more granular and look at tickets created and solved by the hour. Keep in mind that the messaging capacity you define also applies separately to live chats in some circumstances.
Aligning omnichannel routing configuration settings with SLAs
Skills
Incorporating skills into your routing solution can be a great way to ensure the correct agents are dealing with your tickets. Having an agent with the necessary skills and knowledge to resolve a ticket helps improve service quality (as measured by CSAT and average handling time). It is important to carefully balance your use of skills because requiring too many skills on a ticket can cause it to get stuck in a queue waiting for a very small pool of agents to become available. When combining skills with custom omnichannel routing queues, it's important to allocate an adequate number of agents with the skills required by the tickets in the queues they receive work from.
To balance the value of using skills with the risk of tickets having to wait too long to be assigned to an agent, we recommend using the skills timeout and skill priority. When you turn on the skills timeout setting, omnichannel routing drops the optional skills from the routing criteria after a ticket spends a specified amount of time at the front of the queue. Each skill that is dropped from consideration increases the number of agents eligible to receive the ticket and, in turn, increases the likelihood of it being assigned. If you choose to use skills and don't configure the skills timeout, tickets will remain at the front of the queue indefinitely until an agent with all of the matching skills (required and optional) becomes available to receive the ticket.
If you're unsure how many tickets are being assigned to a group with and without skills, you can use the Support data set in Explore to create a report highlighting this information over a specified period of time.
For more information, see About using skills to route tickets.
Auto-accept for messaging
The standard messaging behavior of offering conversations to agents rather than assigning it can increase the amount of time it takes for an end user to receive a first reply. If this is a concern to you, omnichannel routing offers the option to automatically assign messaging conversations to agents, the same way it treats email tickets. We highly recommend using this feature to improve first reply time metrics.
Note that the auto-accept setting can't be used with messaging reassignment timing. You must choose one or the other.
Messaging reassignment time
Messaging reassignment time is another option that can improve the first reply time for messaging conversations. When omnichannel routing is configured to offer messaging tickets to agents rather than assign them directly, this setting ensures the routing engine moves on to the next available agent if it takes an agent too long to accept the ticket. The default messaging reassignment time is 30 seconds, but you can modify this on Enterprise plans. If you have a particularly short reply time SLA for messages, you might want to reduce the message reassignment time.
Configuring idle timeout for unified agent statuses can also help minimize the time messages spent being offered to agents because inactive agents will automatically be set to a status that isn't eligible to receive work from real-time channels such as messaging.
Messaging activity routing
The messaging activity routing setting is a way to automatically free up agent capacity for messaging conversations (and live chats) by automatically setting conversations to inactive after 10 minutes without a response from the end user.
Tickets that become inactive while in the queue are assigned directly to agents rather than offered. This increases the likelihood that tickets with active engagement from end users make it to the front of the queue while the end user is still present and awaiting a response. Likewise, if an end user stops replying to a messaging conversation that is already assigned to an agent, omnichannel routing automatically marks the conversation as inactive and frees up that agent's capacity to be assigned more active messaging tickets.
This standard omnichannel routing behavior can help you achieve your first reply SLAs on the newest messages, but they can also be problematic depending on your organization and workflow. For example, if many messages are received during non-business hours when none of your agents are available, the messaging conversations become inactive and can all be assigned to the first agent to sign on the next morning, overwhelming them with work even while appearing to have spare capacity.
Turning on messaging activity routing causes omnichannel routing to treat all messages (active and inactive) equally, regardless of how old they are. This ensures customer messages are worked in the order they are received, regardless of recent activity. This solves the problems of agents being overwhelmed with inactive tickets and inactive tickets languishing unresolved with an agent who's focused on active conversations, but it also means agents must manage their own messaging capacity by manually changing the status of messaging tickets with unresponsive end users to Pending or Solved.
Focus mode
Turning on focus mode causes omnichannel routing to only assign an agent work from one real-time channel (calls, messaging, or live chat) at a time. In some cases, allowing the agent to focus entirely on one interaction at a time can improve the quality of the interaction and can speed up the resolution. However, it is likely to lengthen your overall first reply time because calls and conversations might remain in the queue longer than they otherwise would.
Other optimizations to support your SLA policies
Omnichannel routing is a powerful tool that can help you achieve your desired service levels. However, sometimes the difference between your actual and desired service levels is so significant that additional staffing is still required. If you've implemented all of these best practices and are still falling short of where you want your service levels, you should reevaluate your staffing and scheduling.