Compendium of Industry benchmarks

Hi Ashley....

I find these benchmarks very useful and agree with most of your estimates. Here are a couple of other benchmarks that could be useful:
a) Percentage of traffic coming from Search Engines (vs promotional / publicity activity on partner sites, using URL, bookmark etc): 25-40%
b) Pages per visit: 7-14
c) Average entries to an online competition: 3,000 - 8,000
d) Ratio to calculate US vs UK performance: 10:1
e) "Viral" / Referral effect of succesful email campaign: 15-25%

For click-thru on banners, it might be useful to note that well placed "entertainment" banners (e.g. film, animation, tv) should achieve a rate of 4-10%

Barney

On 13:55:25 21 March 2003 Ashley wrote:
>A question that is often asked: "What is the industry
>standard for conversion rates / click through rates / drop
>out rates / traffic levels etc. etc.?"
>
>Most companies are only now beginning really to get their
>web metrics house in order but the next obvious question
>to be asked is "Well, I'm now pretty confident I'm
>getting X but is that good or bad?". Over time you
>can benchmark performance relative to your own site but
>what about in comparison to others, in particular your
>peers and competitors?
>
>Is anyone the de facto source of benchmarks for this kind
>of data? Not that I know of. There's no BARB equivalent
>for the web yet. Most information on benchmarks is
>anecdotal.
>
>So what do we all think we believe are benchmarks to guage
>performance by? Following are a few starter suggestions
>(based on an aggregate of all sorts of sources):
>
>- Browse-to-buy ratios on a B2C site: 1-10%
>- Converting registered site users to paid subscribers:
>2-7%
>- Average click through on banners: 0.5%
>- Average click through on e-mail marketing for an
>in-house opt-in list: 10-25%
>- Average click through on e-mail marketing for a
>bought-in targeted opt-in list: 5-15%
>- Percentage of repeat users / unique users for a content
>site: 5-40%

From our sector (e-commerce, mainly B2C) the oft-quoted benchmark for browse to buy is 1-2%

Many quote higher, and our mission is to get people there, but something to bear in mind is the segmentation of this. I believe you also need to consider:

* Conversion of 1st time visitors to buyers

* Conversion of repeat visitors to buyers

* Conversion of returning buyers to additional buyers

Knowing each, and comparing the effects of changes to your site, can mean a heck of a lot when it comes to planning your marketing spend, product mix, pricing etc.

Robin

My page views could be high... these are "entertainment sites", but this is a fairly solid benchmark for sites in this sector - probably simply reflecting the number of interesting sections on our sites (which tend to be similar for all films etc)

Interestingly this figure can be double if "high score" leagues are added to simple games. Not only are more games played, but also this "community" feature - users can see themselves interacting with other users - actually increases people's propensity to browse around the site once they have played.

I'd be interested to see if anyone has any benchmarks for other more classic "community" type activity. I know, for instance, that introducing community boards onto sites increases return visits and increases overall traffic by around 25%.

Barney

On 17:59:28 24 March 2003 Ashley wrote:
>Thanks for those Barney - very interesting.
>
>I would say your average pages per visit was a bit high?
>My benchmark there would be 5-7. Perhaps it is more in
>your sector - all those hugely compelling and engaging
>entertainment sites ;) ?
>
>How can you calculate average entries to a competition?
>How many people have you invited to take part in the first
>place and were they registered/opted-in etc.? Average
>response rates for a compelling competition might be more
>relevant?
>
>US to UK ratio thing is interesting - guess this will vary
>by sector? Or is the online US population 10 X bigger...?
>
>Viral effect also a good one to benchmark.

I think this is all great stuff - but - it’s important that you consider cash in (gross profit) and cash out (operating expenses) alongside these numbers.

Perhaps a few KPI's I could offer are:

Number of new subscribing members - indicates site velocity and appeal
Number of transactions - indicates customer satisfaction
GP per transaction - cash coming in against cost of employed activity
OPEX per transaction - cash going out against cost of employed activity.

These are all primary KPI's under each of these its important to consider secondary drives as outlined below from Alex.

I think you have to be very careful in considering traffic and conversion ratios etc - if the quality of this traffic is not correctly measured you may still be looking at the wrong number to drive business performance.

What do you all think?

Phil

On 10:39:41 27 March 2003 Alex Chudnovsky wrote:
>The following stats are based on about 1,200 email
>campaigns that I was involved with in the last 3 years in
>Jungle.com:
>
>CTR Conversion SalesPerEmail, £
>
>Non-targeted: 2-4% 2-4% 0.03-0.10
>Sales history targeting: 3-8% 4-7% 0.05 - 1.00
>Experimental targeting: 20-40% 10-40% 3.00-10.00
>
>It should be said that in my experience response,
>conversion and sales per email
>
>Definitions we used:
>
>CTR: Total number of clicks from email campaign (not
>unique)
>
>Conversion: number of PLACED order divided by CTR
>
>SalesPerEmail: Total PLACED sales dividing by number of
>recipients on the list.
>
>Experimental targeting: a highly effective targeting
>technique that I developed (invented even?) that is a bit
>too early to tell the whole world about.

That's great - re industry benchmarks & KPI's.

If you were to select 5 top line KEY performance indicators (industry benchmarks) which would they be?

On 13:19:08 1 April 2003 Alex Chudnovsky wrote:
>I'd like to apologize for providing not KPIs but general
>email marketing stats. I'd like to fix my error now.
>
>The KPIs that we used back in Jungle.com were (in no
>particular order):
> 1. Total subscribers
> 2. Transactions (orders), emails vs total on site
> 3. Sales, emails vs total on site
> 4. Average Order Value (AOV) , emails vs site
> 5. Sales per email
> 6. GP per email, GP from all emails vs site
> 7. Drop off ratio (unsubscribers)
> 8. Clicks (total, from text and HTMl parts etc)
> 9. Email opens
>10. Conversion as % of clicks
>11. Conversion as % of the list
>
>All of this was aggregated by various time frames and was
>available on per campaign basis. There were other metrics
>unique to us such as number of customer reviews from
>campaign etc.
>
>Our costs of mailing were 0, hence not provided. I agree
>however that it should be one of the KPIs.
>
>>These are all primary KPI's under each of these its
>>important to consider secondary drives as outlined
>below
>>from Alex.
>
>The most important factor is GP, however margins %-wise
>were generally stable for us (don't think its unusual as
>many businesses have processes in place that ensure
>stability of margins or some growth, not that they always
>work) , so for simplicity's sake we used sales per email
>as primary objective that must be MAXimized.
>
>Secondary objective is to reduce drop off (attrition)
>rate, we want to MINmize it. I forgot to provide stats on
>this one, so here we go: 0.20-0.10% of the list using easy
>to unsubscribe method. This has gone down from 0.60% due
>to improvements in relevance and design.
>
>This all should be viewed in light of MANAGING EXISTING
>CUSTOMERS, I left out CUSTOMER ACQUSITION side
>of things (can cover if you want).
>
>>I think you have to be very careful in considering
>traffic
>>and conversion ratios etc - if the quality of this
>traffic
>>is not correctly measured you may still be looking at
>the
>>wrong number to drive business performance.
>
>I've designed and implemented Jungle's email and affiliate
>tracking system, so I knew exactly how it works and
>ensured that it worked in a way that is usable for
>analysis. Quality of traffic was measured in one uniform
>way, so it was possible to do quality comparisons. I agree
>that one must know how its all workign and be sure numbers
>are comparable, otherwise its easy to make wrong
>decisions.
>
>In our quest to MAXimize primary objective (sales per
>email) we have 2 main variables:
>
>Response rate (CTR) and Conversion (C). Naturally, I need
>to MAXimize both of these to achieve maximum effect, and
>while increasing CTR is fairly easy (fool customer to
>click on these), conversion is the hard bit. What I found
>is that approach that we taken to increase conversion also
>increased response rate as well. Naturally, you've gotto
>be careful about margins -- its easy to generate response
>with high conversion on below cost products.
>
>For a moment I ignored AOV and margins as these were
>pretty much on the same level all the time, there were
>some exceptions of course.
>
>I hope that clarifies and expands my previous post.