CAPTCHAs and conversion rates

17 Feb 2009 - 10:54am
3 years ago
8 replies
861 reads
Harry Brignull
2004

Does anyone have any stats on captchas and the impact they have on
conversion rates?

Marcelo Calbucci of Sampa.com reported that they have a 10% uplift in
conversion rates by removing captchas ( http://tinyurl.com/bqtdef ). Anyone
got any more figures like this?

- Harry

On Tue, Feb 17, 2009 at 3:23 PM, Nikhil Paul <nikhil.paul at itsme.it> wrote:

> Hi Harry,
>
> Thanks for clarifying. In that case its a great idea. In my project blog
> http://www.itsme.it/ well as of now its kinda institutional, but we are
> trying to upgrade the website into a more interactive one, with a blog type
> forum to encourage developer communities and interaction design community to
> get together and discuss about interactive systems etc. Lets see how it
> goes. We plan to get online in the month of march. And i think it will be
> great to direct posts focusing on UX to the collaborative blog site you are
> talking about.
>
> Nikhil
>
>
>
>
>
>
>
> On Tue, Feb 17, 2009 at 4:07 PM, Harry <harrybr at gmail.com> wrote:
>
>> Nikhil,
>>
>> I agree, the quality of the discussions on the IxDA mailing list are
>> outstanding.
>>
>> Just to clarify, this site would not have any formal affiliation with
>> IxDA. It will simply be a site somewhere else on the web where people can
>> add their rss feeds to a 'river of news'. I thought it might be useful to
>> newcomers to the field, people who aren't big RSS-heads, and so on.
>>
>> Harry
>>
>>
>

Comments

17 Feb 2009 - 11:35am
mcaskey
2008

Harry, I don't have any numbers for you, but I will say that my previous
employer's conversion rates did climb significantly when captchas were
removed.

I think the idea of captcha is great, but the implementation usually
stinks. Even the ReCaptcha's aren't very good, and those are the best
around, IMO.

Hard to make a slippery slope when it's covered in trees.

Harry wrote:
> Does anyone have any stats on captchas and the impact they have on
> conversion rates?
>
> Marcelo Calbucci of Sampa.com reported that they have a 10% uplift in
> conversion rates by removing captchas ( http://tinyurl.com/bqtdef ). Anyone
> got any more figures like this?
>
> - Harry
>
>
> On Tue, Feb 17, 2009 at 3:23 PM, Nikhil Paul <nikhil.paul at itsme.it> wrote:
>
>
>> Hi Harry,
>>
>> Thanks for clarifying. In that case its a great idea. In my project blog
>> http://www.itsme.it/ well as of now its kinda institutional, but we are
>> trying to upgrade the website into a more interactive one, with a blog type
>> forum to encourage developer communities and interaction design community to
>> get together and discuss about interactive systems etc. Lets see how it
>> goes. We plan to get online in the month of march. And i think it will be
>> great to direct posts focusing on UX to the collaborative blog site you are
>> talking about.
>>
>> Nikhil
>>
>>
>>
>>
>>
>>
>>
>> On Tue, Feb 17, 2009 at 4:07 PM, Harry <harrybr at gmail.com> wrote:
>>
>>
>>> Nikhil,
>>>
>>> I agree, the quality of the discussions on the IxDA mailing list are
>>> outstanding.
>>>
>>> Just to clarify, this site would not have any formal affiliation with
>>> IxDA. It will simply be a site somewhere else on the web where people can
>>> add their rss feeds to a 'river of news'. I thought it might be useful to
>>> newcomers to the field, people who aren't big RSS-heads, and so on.
>>>
>>> Harry
>>>
>>>
>>>
> ________________________________________________________________
> Welcome to the Interaction Design Association (IxDA)!
> To post to this list ....... discuss at ixda.org
> Unsubscribe ................ http://www.ixda.org/unsubscribe
> List Guidelines ............ http://www.ixda.org/guidelines
> List Help .................. http://www.ixda.org/help
>
>

17 Feb 2009 - 11:48am
Mike Padgett
2008

CAPTCHAs in their normal sense are not typically accessible either, being entirely visual in nature:

http://www.w3.org/TR/turingtest/

Thanks,

Mike

-------------------
www.mikepadgett.com

>
>Harry, I don't have any numbers for you, but I will say that my previous
>employer's conversion rates did climb significantly when captchas were
>removed.
>
>I think the idea of captcha is great, but the implementation usually
>stinks. Even the ReCaptcha's aren't very good, and those are the best
>around, IMO.
>
>Hard to make a slippery slope when it's covered in trees.
>
>
>Harry wrote:
>> Does anyone have any stats on captchas and the impact they have on
>> conversion rates?
>>
>> Marcelo Calbucci of Sampa.com reported that they have a 10% uplift in
>> conversion rates by removing captchas ( http://tinyurl.com/bqtdef ). Anyone
>> got any more figures like this?
>>
>> - Harry
>>
>>
>> On Tue, Feb 17, 2009 at 3:23 PM, Nikhil Paul <nikhil.paul at itsme.it> wrote:
>>
>>
>>> Hi Harry,
>>>
>>> Thanks for clarifying. In that case its a great idea. In my project blog
>>> http://www.itsme.it/ well as of now its kinda institutional, but we are
>>> trying to upgrade the website into a more interactive one, with a blog
>type
>>> forum to encourage developer communities and interaction design community
>to
>>> get together and discuss about interactive systems etc. Lets see how it
>>> goes. We plan to get online in the month of march. And i think it will be
>>> great to direct posts focusing on UX to the collaborative blog site you
>are
>>> talking about.
>>>
>>> Nikhil
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> On Tue, Feb 17, 2009 at 4:07 PM, Harry <harrybr at gmail.com> wrote:
>>>
>>>
>>>> Nikhil,
>>>>
>>>> I agree, the quality of the discussions on the IxDA mailing list are
>>>> outstanding.
>>>>
>>>> Just to clarify, this site would not have any formal affiliation with
>>>> IxDA. It will simply be a site somewhere else on the web where people can
>>>> add their rss feeds to a 'river of news'. I thought it might be useful to
>>>> newcomers to the field, people who aren't big RSS-heads, and so on.
>>>>
>>>> Harry
>>>>
>>>>
>>>>
>> ________________________________________________________________
>> Welcome to the Interaction Design Association (IxDA)!
>> To post to this list ....... discuss at ixda.org
>> Unsubscribe ................ http://www.ixda.org/unsubscribe
>> List Guidelines ............ http://www.ixda.org/guidelines
>> List Help .................. http://www.ixda.org/help
>>
>>
>________________________________________________________________
>Welcome to the Interaction Design Association (IxDA)!
>To post to this list ....... discuss at ixda.org
>Unsubscribe ................ http://www.ixda.org/unsubscribe
>List Guidelines ............ http://www.ixda.org/guidelines
>List Help .................. http://www.ixda.org/help

17 Feb 2009 - 12:01pm
Mark B.
2009

A response from Luis von Ahn of reCAPTCHA:

"Some of our big customers have done A/B tests that show that using
reCAPTCHA does not change conversion rates for them. In general, if
you're being attacked by spam, using a CAPTCHA is a good idea, since
spam DOES decrease usage of your site. On the other hand, for small
sites that have no spam problem, using a CAPTCHA is probably not a
good idea."

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Posted from the new ixda.org
http://www.ixda.org/discuss?post=38813

17 Feb 2009 - 2:07pm
Albert Banks
2009

I agree, we tend to discourage using CAPTCHA in sites we build. There
are a number of coding methods of preventing spam including form
validation, data validation, blacklists, etc. that don't frustrate
users.

Albert
http:www.myjive.net

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Posted from the new ixda.org
http://www.ixda.org/discuss?post=38813

18 Feb 2009 - 4:12am
Harry Brignull
2004

Thanks everyone for their input. Really helpful!

I'm a little unconvinced but the statement "In general, if you're
being attached by spam, using a captcha is a good idea...". It's not
that clear cut - captchas have costs and benefits. It makes sense to
at least start by trying the approaches that are invisible to the user
(i.e. honey pots etc). If the spam problem still proves too great,
then resort to captcha as a backup, rather than as a first response.

Interesting to see that in the sampa.com case study, Marcelo said
their conversion rate went from about 10% to 10.9% conversion. (i.e.
removing the captcha meant that almost one extra person per hundred
completed the activity). This isn't insubstantial, but I'd expected it
to be higher than that, since personally I always struggle with
captchas.

As Jeff said, the other thing to consider is that they aren't a
uniform commodity, some are easier than others; while also user
journeys can vary wildly. If a user spends 10 minutes signing up to a
service that's highly important to them, a simple captcha isn't going
to deter you much. But for a 3 second transient interaction (like a
quick comment or thumbs up), a tricky captcha is going to feel
inappropriately heavyweight.

I'd really hoped to discover a large scale quant study on this... Anyone?

Harry

----
www.90percentofeverything.com

18 Feb 2009 - 4:15am
Harry Brignull
2004

Thanks everyone for their input. Really helpful!

I'm a little unconvinced but the statement "In general, if you're
being attached by spam, using a captcha is a good idea...".

captchas have costs and benefits. It's not that clear cut. It makes
sense to at least start by trying the approaches that are invisible to
the user. If the spam problem still proves too great, then resort to
captcha as a backup, rather than as a first response.

Interesting to see that in the sampa.com case study, Marcelo said
their conversion rate went from about 10% to 10.9% conversion. (i.e.
removing the captcha meant that almost one extra person per hundred
completed the activity). This isn't insubstantial, but I'd expected it
to be higher than that, since personally I always struggle with
captchas.

As Jeff said, the other thing to consider is that they aren't a
uniform commodity, some are easier than others; while also user
journeys can vary wildly. If a user spends 10 minutes signing up to a
service that's highly important to them, a simple captcha isn't going
to deter them much. But for a 3 second transient interaction (like a
quick comment or thumbs up), a tricky captcha is going to feel
inappropriately heavyweight.

I'd really hoped to discover a large scale quant study on this...
Looks like there's nothing publicly available?

Harry

--
www.90percentofeverything.com

18 Feb 2009 - 4:51am
Adrian Howard
2005

On 17 Feb 2009, at 09:01, Mark B. wrote:

> A response from Luis von Ahn of reCAPTCHA:
>
> "Some of our big customers have done A/B tests that show that using
> reCAPTCHA does not change conversion rates for them. In general, if
[snip]

Of course folk visiting "big customers" may be willing to expend more
effort than when they visit other sites.

I know I've struggled through CAPTCHA's for Google/Yahoo coz I
_needed_ a bit of functionality. Other sites I just skipped out the
third time I got it wrong because I'm doing competitive shopping and
the three other stores I'm looking at don't ask me to jump through
that particular hoop...

It's a barrier that doesn't help the user. It just helps the site.
That tells me I should be trying to remove it when ever I can.

Cheers,

Adrian

27 Jan 2011 - 10:23am
gavinwye
2010

Almost a year has passed since the last post on this thred.

In the time that has passed has anyone carried out a detailed study relating to this?

I have come up against this problem twice in the past month and would like to be able to point to some definitive research.

Thanks

Gavin

 

Syndicate content Get the feed