What Gives? Sales 67% down - traffic no change

[ehgee]

Just can't seem to figure it out. There is absolutely no change in traffic (country wise stable / total uniques stable even page-wise / keyword wise stable ) So traffic throttling does not seem to be an issue here; yet sales are down by as much as 67% over the last 3 days.
This is a statistical impossibility seeing the trends on our 12-year old website.
Given the nature of our product, unless we tank in SERPS, nothing, absolutely nothing results in such huge loss of sales. Yet, the SERPS don't show a darn thing.
Any help or hints that anyone can offer would be greatly appreciated?

[scottsonline]

Two things are going on. One is the US economy is collapsing again as fuel prices rise and two google has become cyberdine and is self aware. We are all seeing these huge swings.

For three days as my place of work had slow sales I noticed tons of informational results getting into transactional inquiries. As an example lets say Nike decides to start selling vegetable seeds next week. This current version of google would rank them ahead of the major tree people because after all they're the top brand around. Even if the landing pages for those products were just "call us for a quote" type pages. Big 'brands' can sell whatever they want. It's the amazon/Wikipedia/eBay phenomena now applied all over. These sites must be authorities in the eyes of google.

IMHO the devaluing of links has been a disaster. The result brought down the people buying a billion links which is good but it in turn promoted a lot of garbage. So we are seeing "trusted" brands rising to the top on anything and everything even if it has no relevance to the product being sold. Amazon could sell spaceships starting Monday and they'd rank ahead of nasa

[Whitey]

@ehgee - some questions :

Which region are you in?

Have you looked at your analytics, in particular the funnel in GA?

Is there consistency of "value" in your inventory versus others ?

What are the retailing trends in your region?

Conversion rates are all over the place from what I understand. Not sure if it's a Google thing, an economic environment thing, both or something else at a high level.

At least new auto orders are up in the US which is a good sign. But recovery and purchasing commitment around the western globe is a little weak and volatile. Check out some key verticals on Google Trends [google.com...] [ even though your question isn't about traffic ]

Commitment to purchase may not be as strong as it used to be via Google. A lot more folks have matured to know where to shop now without going via the SERP's, even though it doesn't explain the volatility you express.

[ehgee]

Thanks for your inputs. But what you seem to be suggesting would be more appropriate if it were a long term phenomenon. I am referring to the last 3 days.
@Whitey

Which region are you in?

India, but targeting US traffic

Have you looked at your analytics, in particular the funnel in GA?

Yep.. nothing in there other than showing drastic drop in flow which is borne out by the pathetic sale score

Is there consistency of "value" in your inventory versus others ?

Again, no change to note over the last 3 days

What are the retailing trends in your region?

stable

It's almost as if people collectively decided to hold off purchases for the week :( .... anything in the US to suggest that?

[Web_speed]

It's almost as if people collectively decided to hold off purchases for the week...

You are definitely not seeing things. Myself as well as a few more WebmasterWorld members running eCommerce sites have noticed this exact same weird trend for couple of months now. As if the sales tap is being turned off for a few days (while traffic numbers stay the same+) and then on again followed by spurt of sales in a 24-48 hours window and then tap is off again for the whole week.

The traffic quality during these "off" periods is bad. Lots of mobile devices, wrong geo location, lots of "web previews", very broad long tail keyword matches. Add to that the worst recession in living memory and you get very bad conversions.

[Web_speed]

I'm also starting to suspect that a large part of this traffic (during the "off" periods) may be "prefetch". There is simply no other logical explanation for this "one page view" zombie traffic that suddenly decide to hold off purchasing for a week.

[Sgt_Kickaxe]

Amazon could sell spaceships starting Monday and they'd rank ahead of nasa

That's a sad statement and unfortunately true. It would take a google human evaluator to manually move nasa back ahead of amazon and then god help anyone else trying to enter the spaceships market when some results are fixed.

Google's current setup REQUIRES anti-trust behavior because if they don't play favorites the results are junk. I get the distinct feeling that their "intent engine" which requires a measure of authority is causing real problems with natural rankings based on page value.

[HuskyPup]

A lot more folks have matured to know where to shop now without going via the SERP's, even though it doesn't explain the volatility you express.

I feel this is much more prevalent than many realise, lots of people have their favourite buying sites now, I know I do, in fact I don't even look anywhere else except my usual half dozen sites or so unless it's for something that I have never purchased before and want to get a general "feel" about the widget and pricing levels.

Believe it or not I had never bought anything at Amazon until a month ago and then within a week, not an offer I hasten to add, I required something else and they had precisely what I wanted at the right price.

Yes, it's a bit about brands however I have had, in general, very good buying experiences over the past 15 years or so and found no need to change.

As to ehgee's issue this would not fit his profile history which makes it difficult to explain. I assume it's not a too costly product? Is it aimed at any specific ethnic group or generally for anyone?

My traffic's been down from specific regions this week, China in particular, simply because of the biggest widget fair in my trade.

Then again you may now be experiencing what some others have written about for some time and getting zombie and non-converting traffic thanks to Google.

[Planet13]

@ ehgee

Yet, the SERPS don't show a darn thing.

Just to confirm your statement, you have checked the SERPs and your rankings are the same as before for your targeted keywords?

~~~~

And just to echo what you are saying, on one site of mine traffic has doubled, but the conversion rate has dropped in half. Net effect: same number of sales as this time last year, despite twice the traffic.

[gadget26]

We need a name for this zombie traffic effect. If @ehgee is just now seeing it, and others of us have suffered from it for many weeks, does this mean that the "malaize" is spreading? Has anybody recovered from it?

I've poured over all the stats I can get my hands on and the only thing I can find is that the search queries qoog lets me see are slanted a bit more towards the information seeking type than the price comparison/ready to buy type. But it's so hard to tell since a very high percentage of referrals don't include keywords now. Is it a coincidence that goog put that change (hide most of the keywords) into effect just before we started seeing this effect?

If we could see the hidden keywords, would we see how goog is manipulating the quality of our traffic?

[tedster]

Here's what I haven't seen analyzed: Is the no-change (zombie) traffic coming from the very same keyword sources?

A number of people here say there's no country variation - earlier other reported seeing country variation, so this is different. What is the factor that makes this traffic non-converting?

[Planet13]

Here's what I haven't seen analyzed: Is the no-change (zombie) traffic coming from the very same keyword sources?

Do you mean, are the same keywords driving the now zombie traffic?

If so, it is hard to tell because of the large amount of cloaked keyword referrals.

Also, I don't know if there is a way to easily see in google analytics what the revenue generated by a keyword is. Is there some way to do that?

[ehgee]

@Planet13 yep... Same ranking spots for keywords.

Another thing I noticed, direct traffic nosedived. Now that is strange. Statistically nothing warrants that sudden drop which is fairly stable otherwise. The frustrating part is that I can't seem to figure this direct traffic source. Certainly doesn't seem that too many people would be bookmarking or typing the site page.

[Planet13]

The frustrating part is that I can't seem to figure this direct traffic source.

Could it be from newsletters that were being sent out before but are not being sent out now? I think that is counted as direct traffic (as opposed to referral traffic).

There is of course the possibility that google analytics messed up and was showing referral traffic or search engine traffic as direct traffic. GA does screw up sometimes.

_{I had a site where overnight the bounce rate for the site dropped in half. Then three months later it went back up to where it had traditionally been.}

[netmeg]

Are we just talking about zombie traffic from Google, or zombie traffic in general? Two different things. One of my sites has been barraged by zombie direct over the last three weeks. IF you see a surge in direct traffic, you might want to read this (long) post in the Analytics forum. More and more of us are being hit by some kind of a bot attack, and nobody can figure it out, because it looks exactly like human traffic. [webmasterworld.com...]

Re conversions - if I were having these issues, and my traffic looked normal in every aspect (and I mean *every* aspect - see previous paragraph) then the first thing I would do would be to look at my site with a fine tooth comb.

Check your cart (run test orders through in every browser) and check that there are no glitches with your payment processor. Look for any error logs on the server, and study recent entries. Invest in a live analytics program (even if short term) and watch how people interact with your site. Check pageload times for key pages and your shopping cart.

Obviously, check to make sure you haven't been hacked in some fashion.

For what it's worth, most of the ecommerce sites I oversee are way up this year in the number of orders, but way down in the value of the average sale. That's a tough nut to crack.

[ehgee]

@Planet13
i did that - checked both newsletters as well as the autoresponder mailings... no anomaly there.

I am also beginning to think that GA is messed up often. I see so many anomalies which I always attribute to my own lack of knowledge (& intellect), rather than dare to point fingers at the big G.

@netmeg
thanks a bunch for the link, though in my case it was not a 'spurt' in direct, rather the opposite.
We have been reviewing all our FTPs over those 3 day period and our shopping process too.
FWIW, the conversions came right back to normal over the weekend, so atleast the process flow part seems alright.

[tedster]

I see so many anomalies which I always attribute to my own lack of knowledge

But do you get to the bottom of the anomaly? Or you just move on without understanding it?

Have you checked that your pages are all tagged properly with GA script?

Also note that we do have an Analytics Forum [webmasterworld.com] here where technical discussions of Google Analytics are welcome.

[ehgee]

Actually, I am guilty of just moving on when I can't seem to figure out.
Yes, the pages are tagged alright. Thanks for letting me know about the Analytics Forum... wasn't aware

[claaarky]

We've been seeing this completely erratic sales/conversion pattern since the June 2011 Panda update but the 'spurt' effect (sales happening in batches within a 15-30 minute spell and then nothing for hours before 15-30 minute spurt) seems have become more blatant in the last few months, even to the point where the customer service team are mentioning how everything seems to happen at once. My staff now refer to it as "our turn".

For months I've been checking our position for the keywords that generated the sales immediately after they happen and again the next day to see if anything has changed and I see us in exactly the same position. There's nothing in our analytics to suggest a different type of traffic, like the zombie traffic people are referring to - country, browser, everything is about the same every day.

The only explanation I can think of is that we must receive a quick bump up on various terms, just for a few seconds, maybe as a Google test or maybe as a way of dishing out the converting positions.

On Tuesday last week our conversion rate and average order value doubled and consequently we had our best day since Panda hit in April 2011. This was a very unusual day because it was an all day 'spurt' with sales happening in a more normal pattern, throughout the day. Every other day that week followed the 15-30 minute spurt pattern, with some spurts being below average sale values and some being above average (not usually a mixture).

It's like there are two spurt settings......value and duration. Last Tuesday we must have been on the 'high value' and 'all day' setting.

But as I say, nothing in our analytics or manual ranking checks for the converting terms that shows anything different happening on the good days/periods compared to the bad ones.

[claaarky]

I should add that despite this bizarre spurt effect, our sales totals (number and total value) at the end of the day/week/month are in proportion to our traffic compared to previous years (i.e. half the traffic equates to half the sales).

So for us the erratic pattern is annoying but ultimately our sales are what we would have predicted based on traffic levels compared to previous years.

It could be all down to behaviour patterns of the buying public but the busy and quiet times defy any non-Google explanation most of the time. When your customer service staff who have been dealing with enquiries and sales for years are asking why a 15 minute spell at 2:45 on a Thursday afternoon should be the busiest time of the week, then it's clearly not normal.

[ohno]

India, but targeting US traffic

Ironic, our conversions are also down, started on Friday. We are in the UK selling UK items yet over the weekend we get loads of traffic from all around the world but one country always sticks out-INDIA! Sorry but I don't buy this is an economy thing, Google is screwed up(either intentionally or not). The traffic is not routing to the correct sites, some personal searches back up what our business sites see. This time it's different for us though, traffic IS well down so even the junk is reducing now!

[ohno]

How's this for a swing, when I typed the above we were 40% DOWN on normal vistors, we are now 90% UP! That's comparing the last 4 Mondays. Oh, & prior to the blip I see two visits from Google :-

ww-in-f83.1e100.net MOUNTAIN VIEW
No referring url

we-in-f80.1e100.net MOUNTAIN VIEW
No referring url

Which I am pretty sure is in reaction to a +1 button press!

[backdraft7]

Google needs a new tab at the top of the page titled "Shop Amazon". That would clear a whole lot of space in the default serps.

[theskunk]

once again i see the same trend as ohno for my 10 year old uk ecom site. big traffic spike today, uk visits, but slow/low transactions. i realise its better to look at longer term trends, but what ohno see's I see too. since march 12st we have taken 40% hit on traffic, but today back to our max capped level.

[theskunk]

that shld be since 1st of march traffic down 40%