I’ve owned a successful ecommerce site since before I even knew what affiliate marketing was. One thing i’ve always done is offered extremely limited support. My site has no phone number you can call and only a contact form. The honest reason for this is that I never thought there was much of a need to deal with idiots asking idiotic questions on a daily basis. It always seemed like it would be a waste of time since at one point I worked for an ecommerce site and answered the phone calls and saw first hand how idiotic they were. Don’t get me wrong, we offer excellent service. You get exactly what you order, delivered quickly and we accept all returns. We under-promise and over-deliver on every order and that has lead to a very loyal customer base.
I was ordering something from another website and saw a live help popup. It asked me if i had any questions and indeed I did so I typed my question and got live help. I was impressed with the service and decided to look into the software that was hosting the live help popup. It was a company called Olark. I found it was really easy to install, full of features and cheap so a marketing test was born. I decided to try this live help app on my ecommerce website.
To make sure things were handled appropriately, i dealt with all of the live help chats myself over the course of 3 days. If anyone can sell someone something it’s me so I figured having myself handle the chat’s would give us the best chance of success (for a test, i’d never do it full time).
My Process:
- Got an Olark account and installed the script (very easy)
- I used a function called the “welcome assistant” which initiates a conversation with the website visitor after xx seconds. I chose 60 seconds.
- The “welcome assistant” would send out a random message from a bunch that I entered. Example: Hi, this is {Name}. Can I answer any questions or help you find anything?
- If someone answered the “welcome assistant” i would get the message in my AIM client and chat there.
One of the things that is kind of cool about these chats is that you can actually see what the person is looking at. Just like if you were a salesperson at a store and guide them towards the best product (or the most profitable).
My test results (this isn’t the most scientific of case studies):
Time: 3 Days (Approx 24 hours logged into live chat)
Chats: 16
Increase in Immediate Sales: 0
Observations:
- The vast majority of visitors have no interest in contact by the site, however, some obviously do.
- Most of the people who want to chat are tire kickers aka window shoppers.
- None of the people I spoke to actually bought.
- A few of the people I spoke to were super appreciative of the help I gave and told me they would only buy from us from now on – but they didn’t buy anything after our chat so either they are lying, or will be buying in the future (this is where tracking becomes difficult)
- People who initiate a chat seem to spend a LOT more time on the site then people who don’t.
- People want advice about what they should do or what they should buy.
Overall, i’d consider this test a fail for live chat. Olark is a pretty sweet service though. Maybe it would work better on a higher priced more complicated product or service.
I would like to try another case study in the future where I put the live help on an affiliate landing page and see how it increases conversions.
Have you tested live chat on a site you own? If so, did it lift your conversions?
P.S. My Favorite Live Chat:
Website Visitor: do you ssuck dick
Me: Yes
Website Visitor: i love yo ass
Me: thanks it is pretty sweet
Website Visitor: i was just wondering since we have some cemminstree can i ha’ yo numba showty
Me: 867-5309
Ad Hustler | Subscribe To Ad Hustler
Disclaimer: NO compensation was received for mentioning Olark in this post
Looks like you got a conversion there… To the Rainbow Squadron.
You are popping up on people after a set time, I wonder about those that pop the window when you are trying to leave the site.
@NegBox – It’s something that could be tested but i’d imagine once they are leaving very few would be willing to start a conversation.
What about you make a little twist to popup the chat box when they leave in order to ask them for feedback, the reason why they are leaving, etc…
Then during the chat session you ask them about the product they were interested in and try to convert them as a salesman?
It’s nice when you can make friends like that 🙂
I think it does help with relationship building. I don’t have experience putting one on my sites but have you considered little bit of a longer run? Three days doesn’t seem like an appropriate test. It could have been the time of the month, days of the week, bad luck……
Anyway, thanks for the info though.
Interesting test. I agree with NegBox, that it would be interesting to test popping the window when people leave. I know some ecommerce sites have success with this approach. But it probably isn’t a fit for everyone.
BTW – the chat transcript is a classic.
Interesting test. I never got around to doing any experimenting with live chat, as I have always ignored them on sites I visit. Good to be aware of the service nonetheless.
hahah yeah thats the thing about that chat… you get those morons that dont want to buy just mess around and talk to people on chat …
How perfect is this timing?! We JUST subscribed to a live chat service to add to our ecommerce site!
I’ll be sure to write a follow up post like this when we get some numbers too. Our items are higher priced items (compared to yours) so that should be interesting to see.
Oh… and I LOVE the chat @ the end. Hilarious!
We tried the live chat for a month. Out of 43 chats, we gained 9 new customers. Most were ready to buy and surprised that someone was actually there to answer their questions. But the bulk of the chats were dieting customers. Never had any chats as interesting as yours, though…
Ive also tested this and it didnt work for me.
What it DOES work is something ive discovered from a test, and from that point i started using it everytime.
Its strictly for leadgen.
I know you do a lot of leadgen so this will interest you.
On the form page, where the lead is generated i use actionpopup, its a script that pops up a pop over when the visitors are leaving the site.
I use this exact wording: “What do you want to ask? Type it below and ill send you the answer right away”
and i ask only for email address.
From that point i started to generate an interesting amount of leads that i would’ve lost if i hadnt use this trick.
People filling in that exit form receives an automatic autoresponder where i guide them to fill in my main form and give them some good reasons to do so. At first i used to reply all of them, then i said fuckit, installed autoresponder and still generate more leads.
If you need more info, you have my email.
Ya, the ecommerce site I tried it on was a lower priced product and the live support was absolutely worthless. People were annoyed and simply did not need it for the product.
I am not yet convinced of the value of the shotgun style of proactive chat, have you tried a similar approach just being there to answer questions from the people who initiate with you?
Also, for a more longitudinal study have you taken a look at our google analytics plugin, http://www.olark.com/extensions/google_analytics I added a few customVariables that will let you track live chat segments through your entire funnel (assuming you have goal and ecommerce tracking turned on)
You’ll be able to see if people you chatted with bought during that session, OR if they bought during a subsequent session.
Interesting. I am a consultant for a CRM company and I have setup many live chat systems for large companies. They do see the value in it otherwise they wouldn’t keep the thing up and staff it. It’s a cost savings (cheaper than the phone, agents can do 3+ chats simultaneously) and also a sales tool.
At my previous company, a domain registrar, we did jump through hoops to track sales to chats but it was profitable.
We setup the popup chats (proactively getting in the users face) and also the customer-initiated live chats.
Also, i disagree that the exit pop chats are not effective. You would not see those fake live chat popups are on so many shady merchants’ websites (rebill and other cpa advertisers) if they didn’t give a boost in conversions.
PS – reading chat transcripts are highly entertaining. I’ve read thousands and they are always amusing.
[…] all started a few days ago when we decided to install a chat client on our ecommerce site. Although Adhustler’s experience with chat has not been the best, we still wanted to see how it would do on our […]
[…] all started a few days ago when we decided to install a chat client on our ecommerce site. Although Adhustler’s experience with chat has not been the best, we still wanted to see how it would do on our […]
We ran live chat for a while on our ecommerce site and it was nothing but a time sucker. It actually invited more questions than were necessary AND people started to use it in lieu of a trouble ticket system. Further we found our CSRs spending, in many cases, in excess of 10 minutes per chat trying to help customers who ultimately never bought. Like you, I think this would work better on a site selling a larger ticket item or service.
Statistically, you’ll see 1 extra sale per 100 chats, so I’m not surprised that 16 interactions resulted in zero.
I recommend a virtual agent, especially triggered on abandon. Then you’re not wasting your time and only realizing the benefit of being able to handle the gross numbers.
recently i was searching about this topic and I found that live support does increase conversation rate…
Found a WP plugin that lets me do live chat which is fine because I generally leave it running on the computer as I work through other items.
Haven’t seen any return from it but it’s been maybe 1 1/2 days of using it so far.
I’ll probably let it go for about another week before I decide on it.
Anyway, thanks for your own study – very timely for me 🙂
We use olark for our live chat at PeerFly and people really seem to love it. Olark does provide an awesome service. Easy to implement and easy for everyone to use.
I use it on my car rental website. I have not tested it for any increase in sales yet, however I know it definitely gives the customer a good experience and therefore, more of a reason to tell their friends. I love that.
I will use a AB split test next week to accurately test conversion rate.
Hello There. I found your blog using msn. This is a very well written article. I’ll be sure to bookmark it and come back to read more of Case Study: Does Live Help Increase Conversions? | Ad Hustler . Thanks for the post. I’ll definitely return.
Impressive.You are discussed the so many point of success.I’m fully agree with you.This information is very effective for every student.