Alright,

here is the scenario

I have a client who wants to do an automated post-call survey.

Simplistic Steps:

1. Call comes into IVR
2. IVR randomly selects a caller and plays a "a survey will be conducted at the end of the call" message, flags call.
3. IVR delivers call (as it normally would)
4. Agent completes call and hits "release/disconnect"
5. Call gets picked up and delivered to an automated phone survey (IVR?)

So... allow me to ramble about a bit and see if any of you can offer me any help or insight

How can we do this? Keeping in mind it's currently an unknown system they are using.

I mean...the IVR will lose control of the call as soon as the agent picks up? Is there anyway that the IVR/call routing utility can retain control over that call and re-deliver it once the agent hit's release?

The only thing I can really think of doing is getting the customer to "select 1 if you want a survey to call you back" and then having a system in place that would record their phone # and give them a call back. But how would that system truly know when the call is over? I'd want the customer to be reached within 5 minutes of the call ending...but no point in calling if the customer is still on the call....

*scratches head*

Regardless of whether your IVR can keep control or not, I think you'd have trouble with customers hanging up.

Even if you've told them at the begining of the call that you'd like them to hang on to complete a short survey, most will forget and some just won't be interested.

If it were me, then I'd setup the IVR questionaire and then get agents to transfer the call in once they were done. This would also allow them to ask the customer if they'd be prepared to take part just before it happens, meaning that the customer will remember that they've consented.

If you just want some callers to complete the survey, then you can ask the agent to select which ones. However, this could be of no use if you want a true random sample, of if the survey is to assess how effective the agent was at handling the call. So perhaps you'd need something built into the computer systems to tell the agent when to transfer a call.

Justin: Just read your post in the other topic, namely:

Asking your customer at the end of the call will skew your results a bit, do you really think your agents are going to ask irate callers?

Therefore, it sounds like you would be concerned about the agents selecting which callers to send.

Just posted a similar response to this on another thread.

We are considering using an IVR to survey customers. Agents "cherry picking" respondents to the survey is something that we have thoought long and hard about.

Our current thinking is to promote the feedback mechanism through our other channels, giving them the option to call the survey direct. Identifying themsleves as part of the survey using their unqiue identifier would then enable us to deliver targetted feedback to individual agents.

Yeah...perhaps a "you have been selected for a survey, after this call please dial BLAH"

But it will be tough to get a good percentage really interested in it...

The vendor that we are in discussions with have case studies from one of the big four banks claiming 40%-60% response rates.

To try and answer the original question as to how to technically implenment the scenario...

The normal way to do this is for the IVR to bridge the original call rather than transfer it. The IVR therefore stays on line (conferenced effectively) while the caller speaks to the agent. The agent can then either hit an escape button ('#' for example) or hangup and the IVR will then continue with it's script and go through the survey. This does of course mean that 2 IVR ports are tied up for each of the survey calls.

There are also more sophisticated ways of doing this using desktop softphones and CTI, but this depends on what you already have in place and your budget!!!

You have offered me some good insight, while I personally don't know enough about our IVR and it's capabilitys to say for certain that will work I will look into it.

Thanks!

Justin,
We have recently implemented exactly what you are asking for with a major insurance company. They too tested a system which required the agent to push the call through to the survey, and this resulted in cherry picking of good calls.

We have provided them with a system which automatically offers callers the opportunity to participate in the survey through the IVR and then on completion of call automatically routes through to the survey, with the agent completely unware of whether they have even been offered the chance to participate.

Please feel free to contact me for further information.
Joanne

Hi Joanne,

We have provided them with a system which automatically offers callers the opportunity to participate in the survey through the IVR and then on completion of call automatically routes through to the survey, with the agent completely unware of whether they have even been offered the chance to participate.

Do you have any further information you could post on how you achieved this? One of our goals is to implement post-call surveys, and this is one of the hurdles we're trying to overcome.

Regards,
Michael

Michael - we also developed an automated post-call survey options last year and rolled this out to several of our clients (spanning various sectors).

When left to agent discretion (as expected and in line with everyone above) .. we experienced the 'cherry picking' effect.

To combat this (and sometimes the simplest solutions are the most effective) we reviewed where the call was offered (as part of a script) and varied which teams offered the surveys, and to which call types.

By making the survey offer mandatory to certain teams at certain times and having the agents position the survey as either positive help or a chance to express true feelings, we feel we directly removed the ability to bias responses.

The response was staggering, with many teams averaging 1 in 3 uptakes on the surveys. The resulting MI has proved invaluable in assisting to identify areas to be addresses. The biggest benefit is the ability to monitor and capture dissatisfaction in real time and reroute the call back to a live agent to resolve outstanding issues. We view this as a major boost to customer service.

Without sounding too 'salesy', we provide this as a managed service to our clients and I'd be happy to discuss some of our findings in terms of good/bad experiences etc.

There is some info on our web-site, www.beCogent.com

Regards, Stephen

Thanks for the info, some good reads there but I don't think that specifically what we are looking for. I (at least) and looking for more of a technical solution to transfer these calls to an IVR invisible from the agent.


By making the survey offer mandatory to certain teams at certain times and having the agents position the survey as either positive help or a chance to express true feelings, we feel we directly removed the ability to bias responses.

Even in the scenario you just mentioned you are still only getting "cherry picked" information. At those "certain times" where it is mandatory to offer don't you think your agents would handle the call much better? I wouldn't say you were removing the ability to bias responses....I would say you were empowering and nearly encouraging your agents to bias them...

Currently I am working as a dialer administrator (unison 3.3.2.4)
Can somebody send me the documents related while implementing a davox dialer.
I would really grateful if any one can send the scripting implimentation face...documentations.my mail id rajeshp112@yahoo.com

I completely understand where you're coming from, in terms of the agents performance placing bias on the survey results.

My comment on 'certain times' may be the slightly misleading part; when making the survey mandatory for specific teams, we do this for a period of weeks rather than hours ... so, granted agents may step-up their performance, but if this can be sustained over several weeks .. then everyone wins!! ;-)

If you do find a suitable system that allows the IVR to be invisible to the agent AND gets a high response rate (not just vendors claims), I would genuinely be interested in any info you could share! Our research into this approach yielded mixed success and found that without adequate explanation or positioning behind the reason for the survey, the drop-out/hang-up rate was too high.

Cheers, Stephen

I appreciate your response, and your solution is one we might take a look at. But I'm with Justin here, I'd love to find the technical solution to transfering these callers to the survey app without the agent knowing. I'd imagine we'd experience a much higher hang-up rate than if the agent had offered the survey to the caller, but of course a transparent transfer would negate the "cherry picking" in this case. It's a challenge, either way.

I was just reading your email. Can u tell more about that.

Rather than post too much on this strand, if you can send me your e-mail address I'll forward some information.

stephen.savage@beCogent.com

Hi Stephen,

My email id is alok22_vats@yahoo.com, you can send me email on this email id. I want some more information if you can. Like my call center is very new, it is in the initial stage. We are first person, who are working on the first process of the company. We do not have MIS department over here, who work on the Microsoft excel. I do not have good knowledge about MS - Excel. We need to make a MS-Excel sheet to enter:
1. Agent Productivity Sheet
2. Quality Things

Can u send me these blank excel formats if you can. I would be very thankful to you.

Regards
Alok Mishra