The first step in building your Assisted Funnel is to
determine which teams at your company need to be
involved. Typically, our clients include the following
groups when implementing a System of Engagement:
Functional Area (Sales,
Overview: The functional area is responsible for defining the core and
secondary KPI’s. This team is responsible for defining success and
Best in Class Example: The Assisted Funnel team is truly cross-functional.
Overview: The digital team is responsible for designing the entry points and
overall experience for the Assisted Funnel.
Best in Class Example: Digital teams will work with the analytics team during
the design phase.
Overview: The contact center provides the “human” component of the
Best in Class Example: Contact centers that are completion / conversion
Overview: The IT organization is responsible for implementing the plans and
Best in Class Example: After implementing, allocating one full-time developer
allows for rapid expansion and provides access to staging environments to
everyone involved with the project’s success.
Overview: The analytics team is responsible for ensuring conversion
analytics are properly implemented.
Best in Class Example: Utilize an analytics provider that captures every event
and provide access to everyone on the success team including third-party
Security & Compliance
Overview: Ensure that everything implemented is secure and compliant.
Best in Class Example: Utilize the knowledge vendors bring in working with
other similar clients.
Each team’s responsibilities are important to the overall success of the Assisted Funnel. This is why it is important to ensure that the teams are communicating and on the same page with the same objectives and
schedules. Most companies have at least one project manager and one
technical manager per business unit, and, depending on the number of
business units, some companies will have a business lead and a technical
lead who span across the business units. As a note for companies that don’t have enough project managers: Glia provides consulting services that can fill the gap and assist with communication best practices, including:
The most important aspect to consider before starting the implementation is ensuring all of the cross-functional teams have signed off on the initial core and secondary KPI’s. The teams should also understand the launch timeline and the timeline that the initial KPI’s need to be reached. Best-in-class companies will also ensure everyone on the project team understands the
financial implications that any delay can cause, both in lost potential revenue and additional cost implications.
Using a System of Engagement, there are several ways the agent can offer CoBrowsing and get on the same page with the caller. Without any plugins, both the agent and caller can navigate on the page simultaneously with their own cursor. Instead of “you see that drop down menu below the form?” the agent can say, “let me take you there”.
Implementing an Assisted Funnel can be accomplished in a matter of days or it can be a cross-functional project that takes several weeks. Most ‘best-in-class’ companies launch within 2-4 weeks with an MVP (Minimum Viable Product) approach by using a solution that works out-of-the box with a lightweight deployment.
That said, some companies prefer to wait on launch until after their Assisted Funnel technology is part of their “Core Flow.” A Core Flow implementation features a form that provides an option at the beginning of the user journey that offers an option to “Work with an Advisor through the application process” or “Complete on your own.” This allows visitors to self-select into a guided experience through the Assisted Funnel.
Some other considerations while implementing the assisted funnel are:
Dynamic changing of online flows based on staffing
Targeting visitors though business logic
Analytics and data warehousing
Compliance systems integrations
Telephony system integrations
While the initial implementation
While the initial implementation may finish at launch, your implementation
team should not go their separate ways as the first stage of success is not on go-live. Instead, it is measured by seeing the success of the initial core and secondary KPI’s. It is common that companies find initial KPIs to be easily achievable, however, new stretch goals must be created.
Training is perhaps the most important part of getting maximum value from the Assisted Funnel, but it is not only about the ability to train your agents on your website and the platform. The most important aspect is training your agents to do the right thing at the right time in real-time, while assisting a customer with an engagement.
Most companies start their assisted conversion programs by taking their
online chat agents and moving them to an engagement platform, allowing for more context and better communication. That is OK to start, but this doesn’t set them up with the best chance to produce significant results. The fundamental issue with online chat agents is that they are inadequately trained and compensated.
For sales use cases, the agents should be recruited and/or trained as sales professionals. They should understand both product value statements, as well as basic sales methodologies. Best practices are essential, and it’s often the simple things that make all the difference. For example, when a customer asks a question, your salesperson shouldn't simply answer the
question and ask: “Is there anything else I can help you with today?” Agents ideally must answer the question and lead in with another.
Here are a few examples:
For service use cases, the agents should not only act as support, but also true teachers. Service agents have the opportunity to show customers, rather than tell them, how to complete an online funnel. This resolves the current concern, while teaching them how to do something themselves for next time.
This solves for Next Issue Avoidance, decreases cost ,and increases
MORTGAGE FUNNEL CASE STUDY
A company we worked with in the mortgage space implemented our engagement platform and started out by taking 25 of their chat agents and moving them to the Glia platform. The agents began leveraging Glia to engage their customers and conversions went up, but only moderately.
After their first transcript review, Glia’s Blueprint Team found that the agents would start an engagement, answer the customer's question, ask if
they had any more, and then let them know they could reach out again if they had another question.
The client then went back and gave a few simple instructions to all of her agents. She said:
1. Do not ever end an engagement (hang up) on a customer
2. Answer the customer’s question by CoBrowsing the customer to the answer
3. After answering the customer’s question, offer to complete the rest of the funnel with the customer right up to the last step
After implementing these three simple rules, the results spoke for themselves, with funnel completion rates increasing by 4X. Later the client continued to add and refine a few more instructions to further increase
completion, especially encouraging the agents to upgrade the conversation
to audio, and even video. Again, the completion rates continued to rise. Notably, the side effect of this increased engagement was that while the
completion rates were increasing, so was customer satisfaction. Not once
was an agent called out as being too overbearing or pushy because the agent’s instructions were not directly conversion related.
Incentive compensation can also be used for both sales and service agents to drive additional impact to the core and secondary KPI’s. Some of the best companies we work with take advantage of an innovative compensation model that our Blueprint team developed. The model is based on weighting the compensation not solely on the end result, but on actions that increase
customer satisfaction, decrease customer effort, and ultimately (based on analytics), increase the potential of achieving the funnel conversion. These actions can be never ending the engagement, using CoBrowsing, upgrading to audio/video or something that is very specific to the funnel, such as ensuring to offer a specific discount.
You can’t effectively improve what you can’t measure. However, after
working with hundreds of companies we have found that measuring
everything leads to companies ending up in a state of analysis paralysis. The key is to isolate the core and secondary KPI’s ,and then create reports with placeholder data. Make sure that each data point maps back to the KPI’s, and if any data points don’t directly match back, they should be removed from the start. While it can’t be said that having too much information is a bad thing, a strong focus on the right information is important when there can be the temptation to over-analyze and get distracted.
Common KPIs we see include:
Average Order Value
Customer Effort Score
Average Handle Time
Many companies also use KPI’s to compensate their agents so it is crucial to have the right reporting in place. Companies also need to be able to determine which agents are outperforming their peers so that great agents can be rewarded. Underperforming agents can be retrained by the best agents, but if that does not work out, they can be moved off the Assisted Funnel teams.
Most companies begin their assisted funnel by driving self-serve traffic
through marketing campaigns and advertising into assisted funnels through a light out-of-the box implementation of an engagement technology or a simple chat function. This is typically a persistent bubble on the website that says something like “Talk to an expert,” an image of a person, or a chat icon. Companies tend to focus on the percentage of their traffic that clicks their button, but we’ve found that they’re looking in the wrong place.
Instead, enterprises should first take a look at their available staff. They
should first think about their total productive time and then the cost of their productive time. From there, they should target the maximization of their productive time.
Most of the time the math still works even if a call center employee’s productivity is as low as 50%. That said, it is ideal to ensure each employee is as productive as possible, thus maximizing returns. With this in mind, technology
reenters the picture.
The new aim becomes pushing as many customers through the assisted funnels as possible, right up to the point where our staff become 100% productive. Regarding an engagement button, image, or icon, experience tells us that we can expect around 2% of visitors to reach out and engage. 1-5% may be enough to start out with, but this is not going to have a material impact on site-wide macro conversion rates.
As companies get more comfortable with the unit economics, they will
allocate more staff and need ways to drive more engagement through their
Assisted Funnels. The next thing they can implement is automated business logic that presents nudges and/or engagement offers when customers attempt certain actions in the funnel that indicate that they are high-value or have high-intent to proceed. This can increase the percentage of assisted customers by up to 7% if done properly.
However, to ensure the largest impact, Assisted Funnel conversion has to be incorporated into the core flow of every business. This means that when a customer is about to start a process online, they have two options:
1. Start the process using self-serve, or
2. Start it with assistance.
Self-Serve vs. Expert?
#2.0 With assistance
#2.1 With assistance
Of course companies can offer different variations of this approach, only targeting higher value customers while considering real-time staffing rates. However, this measure can bring the percentage of assisted customers to over 15% if implemented right. This is when companies are able to begin seeing a material impact on site-wide macro conversion rates.
The second area for optimization is increasing the conversion rates of the visitors who do engage with the assisted channel. This can be accomplished by better agent training, optimization of incentive compensation structures, or through an automated artificial intelligence engine that can assist/manage
your agents in real-time.
Agent training can be optimized through:
• Engagement reviews: Historical look back at randomized engagements with a manager to point out areas of improvement.
• Mystery shopping: Live professionally done mystery shopping where the agent is challenged on specific instructions they are given during training.
• Peer-to-peer training: Having the top agents teach the rest of the agents what they are doing to be successful.
Incentive compensation can be optimized through changing the compensation models to account for specific actions that drive conversion. For example, if a company learns that CoBrowsing increases the chance of conversion
by 60%, the incentive compensation that is offered can be to incentivize
CoBrowsing. Companies can pay on the conversion event, but by paying on elevated customer experience, companies are able to drive an increase, in not only conversion, but also customer satisfaction or NPS.
Perhaps one of the most effective ways to optimize a company's agents is to use an artificial intelligence engine, such as Glia’s OmniGuide, which
assists the agent in real-time, offering suggestions and prompts to the agent based on what the customer is saying. For example, if a customer asks, “how much life insurance should I buy?” the AI engine reads what was typed and
then suggests to the agent “Most of our clients purchase 10X of their annual salary”. The AI engine can also assist with cross-selling, up-selling, agent accuracy, and fraud detection.