This is default featured slide 1 title

This is default featured slide 1 title

You can completely customize the featured slides from the theme theme options page. You can also easily hide the slider from certain part of your site like: categories, tags, archives etc. More »

This is default featured slide 2 title

This is default featured slide 2 title

You can completely customize the featured slides from the theme theme options page. You can also easily hide the slider from certain part of your site like: categories, tags, archives etc. More »

This is default featured slide 3 title

This is default featured slide 3 title

You can completely customize the featured slides from the theme theme options page. You can also easily hide the slider from certain part of your site like: categories, tags, archives etc. More »

This is default featured slide 4 title

This is default featured slide 4 title

You can completely customize the featured slides from the theme theme options page. You can also easily hide the slider from certain part of your site like: categories, tags, archives etc. More »

This is default featured slide 5 title

This is default featured slide 5 title

You can completely customize the featured slides from the theme theme options page. You can also easily hide the slider from certain part of your site like: categories, tags, archives etc. More »


With a Whiff, Mice Can Transmit Pain to Each Other

What hurts one mouse, hurts every mouse.

That’s the conclusion of a new study examining the social transfer of pain in mice. When one group of mice was exposed to a painful stimulus, a completely unaffected group displayed the same kind of heightened sensitivity as the first. Given that mice are mammals like us, the effect could also exist in humans, as well as informing future pain research.
Testing for Pain
In their study, researchers from the Oregon Health and Science University work

Original Source File

The 7 Epiphanies Needed to Intuitively Grasp Statistical Significance

There is only one danger more deadly to an online marketer than ignorance, and that danger is misplaced confidence.

Whenever a marketer omits regular statistical significance testing, they risk infecting their campaigns with dubious conclusions that may later mislead them. But because these conclusions were based on “facts” that the marketer “empirically” observed with their own two eyes, there is scant possibility that these erroneous ideas will ever be revisited, less questioned.

The continued esteem given to these questionable conclusions causes otherwise sane marketers to irrationally believe their sterile photos to be superior, their shoddy headlines to be superb, and their so-so branding to be sublime.

Statistical significance testing is the cure to this woe. There are math courses aplenty that describe this field, but the world has no need for another. Today, I propose something different: a crash course in the necessary intuition. The goal of this piece, then, is to instill in you a series of “a-ha” moments that’ll make statistical significance click all while sending warm, fuzzy rushes of understanding into your mind.

Epiphany #1: Large sample sizes dilute eccentricity

Imagine you see a book rated an “average” of 5 stars on Amazon. If this average was based on the review of only a single reader, you would hardly think this book better than another which was rated lower (say 4.2 stars), albeit on the back of hundreds of reviews. Common sense informs you that the book with one rating might have been reviewed by a reader who, for purely idiosyncratic reasons, happened to adore it. But you, as a cautious potential purchaser, cannot tell whether that single review was more reflective of the reviewer rather than the book. Without further information, you cannot be confident of the book’s quality.

As you can see with this Amazon book review example, small sample sizes give eccentricity a chance to express itself. It’s for this exact same reason that an advertising campaign report containing only five clicks makes for an unreliable source of truth. Here, it’s possible that those five advert-clickers were oddly passionate fans of your product who happened to see your advert at the right time. The excellent results your advertising campaign seemed to enjoy may, in reality, have just been a fluke.

The intuitive idea we have just seen has not gone unnoticed by mathematicians. They have indeed packaged it up with a delightfully self-explanatory name: “The Law of Large Numbers”.

Epiphany #2: P-values are trade-offs between certainty and experiment length

Imagine a gambler who bets his house on a coin being rigged. We would pity his stupidity if he bet after seeing a coin land “heads” only three times in a row. But no one would doubt this mental acuity if the coin had instead landed “heads” a million times in a row. Intuitively—and indeed mathematically—this is a sound bet.

But notice that the gambler is still betting. He can never be fully, totally, and absolutely certain that the coin is rigged. Even after seeing a million “heads” in a row, there is still an infinitesimal yet nevertheless existent chance that a fair coin could have given the result of “a million heads in a row”. But, practically speaking, this is exceedingly unlikely, so the gambler shouldn’t let such a tiny shard of uncertainty deter him from making a fundamentally sound bet.

Now we have two extremes: flipping a coin three times, after which it is too early to make a confident bet; and flipping a coin a million times, after which it is exceedingly secure to bet. But what if our gambling man has a family wedding to attend that afternoon. He still wants to bet on the coin in confidence, but he doesn’t want to wait around until it has been flipped “heads” a million times in a row. Translated into the business context, what if we, as advertisers, don’t want to continue our Puppies vs Kitten Photo A/B test for 10 years before deciding which photo was better for sales? By waiting that long, we would have wasted 10 years in showing a proportion of our customers a photo that was comparatively ineffective at effecting sales. Had we figured out which photo was better earlier on, we could have had our best foot forward for a much longer period and earned higher profits all that while.

The crux of the matter is this: There is a trade-off between certainty and experiment length. We can see this intuitively by considering how our feelings of confidence would develop after an increasingly long series of coin flips. After 1 flip of “heads”, none of us would suspect the coin of being rigged. After 5 flips, we’d start seriously entertaining the thought. After 10 flips, most of us would strongly suspect it, but perhaps not enough to bet the house on it. After 100 flips, little doubt could remain in our minds, and we’d feel confident about making a serious wager. After 1,000 flips, we’d be screaming at the top of our lungs for the bookie to take our money.

As we have seen, the more consecutive “heads” flips we witnessed landing, the more certain we’d feel about the coin being rigged. But given that we are not immortal and that we will never reach 100% certainty with anything in our lives, we all must choose a pragmatic point where our uncertainty reaches a tolerably low level, a point where we put our hands up and say, “I’ve seen enough—let’s do this thing”.

This trade-off point is quantified by statisticians with a figure they dub the p-value. Very roughly speaking, the p-value corresponds to the chance you have of being wrong about your conclusion. The p-value can thus be thought of as a preference, one that represents your desired trade-off between certainty and experiment length. Typically, marketers set their p-value to .05, which corresponds to having a 1 in 20 chance of being wrong. If you are risk averse about making mistakes, you could set your p-value to .01, which would mean you have only a 1 in 100 chance of being wrong (but your experiment would take much longer to attain this heightened level of certainty).

jelly-beans-statisical-significance(Image Source)

Perhaps no industry is as wedded to the use of p-values as the pharmaceutical industry. As Ben Goldacre points out in his chilling book, Bad Pharma, there is terrible potential for the pharmaceutical industry to hoodwink doctors and patients with p-values. For example, a p-value of .05 means that one trial in 20 will incorrectly show a drug to be effective, even though, in actuality, that drug is no better than placebo. A dodgy pharmaceutical company could theoretically perform 20 trials of such a drug, bury the 19 trials showing it to be rubbish, and then proudly publish the one and only study that “proves” the drug works.

For the same probabilistic reasons, the online marketer who trawls through their Google AdWords/Facebook Ads/Google Analytics reports looking for patterns runs a big risk of detecting trends and tendencies which don’t really exist. Every time said marketer filters their data one way or the other, they are essentially running an experiment. By sheer force of random chance, there will inevitably be anomalies, anomalies which the marketer will then falsely attribute to underlying pattern. But these anomalies are often no more special than seeing a coin land “heads” five times in a row in 1/100 different experiments where you flipped five fair coins.

Epiphany #3: Small differences in conversion rates are near impossible to detect. Large ones, trivial.

Imagine we observed the following advertising results:


Upon eyeballing the data, we see that the goat variant tripled its equestrian competitor’s conversion rate. What’s more, we see that there was a large number of impressions (1,000) in each arm of the experiment. Is this enough to satisfy the aforementioned “Law of Large Numbers” and give us the certainty we need? Surely these data mean that the “Miniature Goat” is the better photo in a statistically significant way?

Not quite. Without going too deep into the math, these results fail to reach statistical significance (where p=.05). If we concluded that the goat was the better photo, we would have a 1 in 6 chance of being wrong. Our failure to reach statistical significance despite the large number of impressions shows us that impressions alone are insufficient in our quest for statistically significant results. This might surprise you. After all, if you saw a coin land “heads” 1,000 times in a row, you’d feel damn confident that it was rigged. The math of statistical significance supports this feeling—your chances of being wrong in calling this coin rigged would be about 1 in 1,000,000,000,000,000,000,000,000,000,000… (etc.)

So why is it that the coin was statistically significant after 1,000 flips but the advert wasn’t after 1,000 impressions? What explains this difference?

Before answering this question, I’d like to bring up a scary example that you’ve probably already encountered in the news: Does the use of a mobile phone increase the risk of malignant brain tumors? This is a fiendishly difficult question for researchers to answer, because the incidence of brain tumors in the general population is (mercifully) tiny to start off with (about 7 in 100,000). This low base incidence means that experimenters need to include absolutely epic numbers of people in order to detect even a modestly increased cancer risk (e.g., to detect that mobile phones double the tumor incidence to 14 cases per 100,000).

Suppose that we are brain cancer researchers. If our experiment only sampled 100 or even 1,000 people, then both the mobile-phone-using and the non-mobile-phone-using groups would probably contain 0 incidences of brain tumors. Given the tiny base rate, these sample sizes are both too small to give us even a modicum of information. Now suppose that we sampled 15,000 mobile phone users and 15,000 non-users (good luck finding those).

At the end of this experiment, we might count two cases of malignant brain cancer in the mobile-phone-using group and one case in the non-mobile-using group. A simpleton’s reading of these results would conclude that the incidence of cancer (or the “morbid conversion rate”) with mobile phone users is double that of non-mobile-phone users. But you and I know better, because intuitively this feels like too rash a conclusion—after all, it’s not that difficult to imagine that the additional tumor victim in the mobile-phone-using group turned up there merely by random chance. (And indeed, the math backs this up: this result is not statistically significant at p=.05; we’d have to increase the sample size a whopping 8 times before we could detect this difference.)

Let’s return to our coin-flipping example. Here we only considered two outcomes—that the coin was either fair (50% of the time it lands “heads”) or fully biased to “heads” (100% of the time it lands “heads”). Phrasing the same possibilities in terms of conversion rates (where “heads” counts as a conversion), the fair coin has a 50% conversion rate, whereas the biased coin has a 100% conversion rate. The absolute difference between these two conversion rates is 50% (100% – 50% = 50%). That’s stonking huge! For comparison’s sake, the (reported) difference between the miniature pony and miniature goat photo variants (from the example at the start of this section) was only .2%, and the suspected increase in cancer risk for mobile phone users was .01%.

Now we get to the point: It is easier to detect large differences in conversion rates. They display statistical significance “early” (i.e., after fewer flips or fewer impressions, or in studies relying on smaller sample sizes). To see why, imagine an alternative experiment where we tested a fair coin against one ever so slightly biased to “heads” (e.g., one that lands “heads” 51% of the time). This would require many, many coin flips before we would notice the slight tendency towards heads. After 100 flips we would expect to see 50 “heads” with a fair coin and 51 “heads” with the rigged one, but that extra “heads” could easily happen by random chance alone. We’d need about 15,000 flips to detect this difference in conversion rates with statistical significance. By contrast, imagine detecting the difference between a coin biased 0% to “heads” (i.e., always lands “tails”) and one biased 100% to “heads” (in other words, imagine detecting a 100% difference in conversion rates). After 10 coin flips we would notice that the results would be either ALL heads or ALL tails. Would there really be much point in continuing to flip 90 more times? No, there would not.

This brings us to our next point, which is really just a corollary of the above: Small differences in conversion rates are near impossible to detect. The easiest way to understand this point is to consider what happens when we compare the results of two experimental variants with identical conversion rates: After a thousand, a million, or even a trillion impressions, you still won’t be able to detect a difference in conversion rates, for the simple reason that there is none!

Bradd Libby, of Search Engine Land, calculated the rough number of impressions necessary in each arm of an experiment to reach statistical significance. He then reran this calculation for various different click-through rate (CTR) differences, showing that the smaller the expected conversion rate difference, the harder it is to detect.


Notice how in the final row an infinite number of impressions are needed; as we said above, we will never detect a difference, because there is none to detect. The consequence of all this is that it’s not worth your time, as a marketer, to pursue tiny expected gains; instead, you’d be better off going for a big win that you have a chance of actually noticing.

Epiphany #4: You destroy a test’s validity by pulling the plug before its preordained test-duration has passed

Anyone wedded to statistical rigor ought to think twice about shutting down an experiment after perceiving what appears to be initial promise or looming disaster.
Medical researchers, with heartstrings tugged by moral compassion, wish that every cancer sufferer in a trial could receive what’s shaping up to be the better cure—notwithstanding that the supposed superiority of this cure has yet to be established with anything approaching statistical significance. But this sort of rash compassion can have terrible consequences, as happened in the history of cancer treatment. For far too long, surgeons subjected women to a horrifically painful and disfiguring procedure known as the ‘radical mastectomy‘. Hoping to remove all traces of cancer, doctors removed the chest wall and all axillary lymph nodes, along with the cancer-carrying breast; it later transpired that removing all this extra tissue brought no benefit whatsoever.

Generally speaking, we should not prematurely act upon the results of our tests. The earlier stages of an experiment are unstable. During this time, results may drift in and out of statistical significance. For all you know, two more impressions could cause a previous designation of “statistically significant” to be whisked out from under your feet. Moreover, statistical trends can completely switch direction during their run-up to stability. If you peep at results early instead of waiting until an experiment runs its course, you might leave with a conclusion completely at odds with reality.

For this reason, it’s best practice not to peek at an experiment until it has run its course—this being defined in terms of a predetermined number of impressions or a preordained length of time (e.g., after 10,000 impressions or two weeks). It is crucial that these goalposts be established before starting your experiment. If you accidentally happen to view your results before these points have been passed, resist the urge to act upon what you see or even to designate these premature observations as “facts” in your own mind.

Epiphany #5: “Relative” improvement matters, not “absolute” improvement

Look at the following table of data:


After applying a statistical significance test, we would see that the 80s rocker photo outperforms the 60s hippy photo in a statistically significant way. (The numerical details aren’t relevant for my point so I’ve left them out.) But we need to be careful about what business benefit these results imply, lest we misinterpret our findings.

Our first instinct upon seeing the above data would be to interpret it as proving that the 80s rocker photo converted at a 16% higher rate than the 60s hippy photo, where 16% is the difference by subtraction between the two conversion rates (30% – 14% = 16%).

But calculating the conversion rate difference as an absolute change (rather than a relative change) would lead us to understate the magnitude of the improvement. In fact, if your business achieved the above results, a switch from the incumbent 60s hippy pic to the new 80s rocker pic would cause you to more than double your number of conversions, and, all things being equal, you would, as a result, also double your revenue. (Specifically, you would have a 114% improvement, which I calculated by dividing the improvement in conversion rates, 16%, by the old conversion rate, 14%.) Because relative changes in conversion rates are what matter most to our businesses, we should convert absolute changes to relative ones, then seek out the optimizations that provide the greatest improvements in these impactful terms.

Epiphany #6: “Statistically insignificant” does not imply that the opposite result is true

What exactly does it mean when some result is statistically insignificant? The example below has a p-value of approximately .15 for the claim that the Mini Goat photo is superior, making such a conclusion statistically insignificant.


Does the lack of statistical significance imply that there is a full reversal of what we have observed? In other words, does the statistical insignificance mean that the “Miniature Pony” variant is, despite its lower recorded conversion rate, actually better at converting than the “Miniature Goat” variant?

No, it does not—not in any way.

All that the failure to find statistical significance says here is that we cannot be confident that the goat variant is better than the pony one. In fact, our best guess is that the goat is better. Based on the data we’ve observed so far, there is an approximately 85% chance that this claim is true (1 minus the p-value, .15 = .85). The issue is that we cannot be confident of this claim’s truth to the degree dictated by our chosen p-value—to the minimum level of certainty we wanted to have.

One way to intuitively understand this idea is to think of any recorded conversion rate as having its own margin of error. The pony variant was recorded as having a .1% conversion rate in our experiment, but its confidence interval might be (using made-up figures for clarity) .06% above or below this recorded rate (i.e., the true conversion rate value would be between .04% and .16%). Similarly, the confidence interval of the goat variant might be .15% above or below the recorded .3% (i.e., the true value would be between .15% and .45%). Given these margins of error, there exists the possibility that the pony’s true conversion rate would be at the high end (.16%) of its margin of error, whereas the goat’s true conversion rate would lie at its low end (.15%). This would cause a reversal in our conclusions, with the pony outperforming the goat. But in order for this reversal to happen, we would have had to take the most extreme possible values for our margins of error—and in opposite directions to boot. In reality, these extreme values would be fairly unlikely to turn up, which is why we say that it’s more likely that goat photo is better.

Epiphany #7: Any tests that are run consecutively rather than in parallel will give bogus results

Statistical significance requires that our samples (observations) be randomized such that they fairly represent the underlying reality. Imagine walking into a Republican convention and polling the attendees about who they will vote for in the next US presidential election. Near everyone in attendance is going to say “the Republican candidate”. But it’s self-evident that the views of the people in that convention are hardly reflective of America as a whole. More abstractly, you could say that your sample doesn’t reflect the overall group you are studying. The way around this conundrum is randomization in choosing your sample. In our example above, the experimenter should have polled a much broader section of American society (e.g., by questioning people on the street or by polling people listed in the telephone directory.) This would cause the idiosyncrasies in voting patterns to even out.

If you ever catch yourself comparing the results of two advertising campaigns that ran one after the other (e.g., on consecutive days/weeks/months), stop right now. This is a really really bad idea, one that will drain every last ounce of statistical validity from your analyses. This is because your experiments are no longer randomly sampling. Following this experimental procedure is the logical equivalent of extrapolating America’s political preferences after only asking attendees of a Republican convention.

To see why, imagine you are a gift card retailer who observed that 4,000% as many people bought Christmas cards the week before Christmas compared to the week after. You would be a fool if you concluded that the dramatic difference in conversion rates between these two periods was because the dog photo you advertised with during the week preceding Christmas was 40 times better at converting than the cat photo used the following week. The real reason for the staggering difference is that people only buy Christmas cards before Christmas.

Put more generally, commercial markets contain periodic variation—ranging in granularity from full-blown seasonality to specific weekday or time of day shopping preferences. These periodic forces can sometimes fully account for observed differences in conversion rates between two consecutively run advertising campaigns, as happened with the Christmas card example above. The most reliable way to insulate against such contamination is to run your test variants at the same time as one another, as opposed to consecutively. This is the only way to ensure a fair fight and generate the data necessary to answer the question ‘which advert variant is superior?’ As far as implementation details go, you can stick your various variants into an A/B testing framework. This will randomly display your different ads, and once the experiment ends you simply tally up the results.

Perhaps you are thinking, “My market isn’t affected by seasonality, so none of this applies to me”. I strongly doubt that you are immune to seasonality, but for argument’s sake let’s assume your conviction is correct. In this case, I would still argue that you have a blind spot in that you are underestimating the temporally varying effect of competition. There is no way for you to predict whether your competitors will switch on adverts for a massive sale during one week only to turn them off during the next, thereby skewing the hell out of your results. The only way to protect yourself against this (and other) time-dependent contaminants is to run your variants in parallel.


Having been enlightened by the seven big epiphanies for understanding statistical significance, you should now be better equipped to pull up your sleeves and dig into statistical significance testing from a place of comfortable understanding. Your days of opening up Google AdWords reports and trawling for results are over; instead, you methodically set up parallel experiments, let them run their course, choose your desired trade-off for certainty vs. experimental time, give adequate sample sizes for your expected conversion rate differences, and calculate business impact in terms of relative revenue differences. You will no longer be fooled by randomness.

About the Author: Jack Kinsella, author of Entreprenerd: Marketing for Programmers.

Original Source File

BEATRICE.B Spring Summer 2017 | CPM Moscow by Fashion Channel

Source Article

With three presidential debates in the can, the loser is…

Earth was hardly mentioned, and that means the loser is actually . . . us

The pundits will no doubt be yammering on for days about who won and who lost the final U.S. presidential debate.

But I’d say that we don’t need the pundits to tell us who the overall loser was. I think it was humanity.

During the debates, we heard a lot of shouting, but almost nothing about the most profound, long-term issue all of us confront: How 7 billion of us, probably growing to 9 billion in relatively sh

Original Source File

2 A/B Testing Failures and 1 Win: Why A/B Testing Failures Are as Important as the Wins

You’ve probably seen a lot of blog posts floating around the internet about A/B testing successes. This blog is no exception.

But hardly anyone talks about their failed experiments. I don’t blame them–it’s hard to expose to the world that you weren’t right about something.

Which leads us to believe…is anyone running failed, insignificant tests? I mean, if no one’s talking about it, it must not be happening, right?

Let me tell you a secret: Everyone is failing. Everyone is running or has run an experiment that got them nowhere.

However, at Kissmetrics, failure is part of our A/B testing process. If none of our tests fail, we know we’re not running enough tests or our ideas are too safe.

In fact, the bigger the failures, the closer we are to an even bigger win.

We’re never 100% correct about our hypotheses. No matter how many years of experience you have, no matter how much you think you understand your customers…there’s always room for learning.

Otherwise, why would we test in the first place?

Now let’s take a look at a couple of our own failures so you can see what I mean.

Failure #1: Too much white space on our product page

Test Hypothesis: There’s too much clutter at the top of the page. By removing the background image and reducing white space, we’ll make the page copy more visible, enticing people to scroll down and interact more with the page.

You already know that this test failed, but just from looking at the hypothesis–do you know why?

I’ll give you a hint: it has a lot to do with data.

We technically had data that indicated a dip in conversion on this page.



However, we didn’t have evidence that people weren’t scrolling down, or that the space at the top was stopping them from converting. When a hypothesis has little or no evidence, we have a slim chance of winning the test.


Improvement over original: 4.41%
Certainty: 55.27% for the variant

What we learned:

No significant data here.

Having a hero image or not in this page won’t influence our conversions. In previous homepage tests, our hero image mattered, but perhaps not on this exact page.

In previous tests on this particular page, we experimented with the copy and overall messaging. Therefore, our next test should be around the copy to see if we’ll get a lift.

Failure #2: Copy and images on the product page

Test Hypothesis: A more benefits oriented product landing page will lead to better quality leads.




Improvement over original: -13.68%
Certainty: 80.12%

Improvement over original: 8.77%
Certainty: 60.78%

What we learned:

In the last test (failure #1), we didn’t change enough of the page for there to be significant results.

This time, we changed 1) the copy, 2) product screenshots, and 3) the overall layout of the page below the hero image.

That’s a lot, right?

Here’s the thing: when we change both copy and design, it’s hard to tell whether it was the copy or the design was the reason for a lift or a decrease. We can’t isolate the variable and know what was responsible for the outcome.

In previous tests, tested the copy first, then test the design after. That’s what we’ll do for our next test.

The Win: Product page headline

Test Hypothesis: Adding a benefit centric headline to the product page will increase demo requests and signups because we’re showing them the the value they’ll get from Kissmetrics. All our happy customers have said in interviews they love seeing individual user behavior in Kissmetrics. But we’ll take that one step further and add the result to the end of the headline.




Improvement over original: 163.46%
Certainty: 97.33%

Requested Demo:
Improvement over original: 507.97%
Certainty: 99.67%

What we learned:

Finally, a win! And it only took us 2 failed tests to get there. Not bad.

The increase in requested demos is huge. We didn’t see 99% significance on the signups, so we can’t say for sure that’s a win–but a 507.97% in demos is worth launching.

The major learning here is that the headline on our product page carries a lot of weight. We didn’t change the rest of the page, or the Call to Action copy. A good next test would be to test the rest of the landing page copy to see how much weight it carries.

And finally, having user interviews and user reviews made our hypothesis strong. Yes, benefit-centric copy is a good thing, but what benefit? What do our ideal customers absolutely love about our product?

Having this research evidence from our customers made the win even bigger.


All failures lead to a big win when you’re extracting a major learning from each test. The best part about these learnings is that they’re unique to YOU and your company–meaning, one test that worked for us might not work for you.

However, the best way to get these learnings and big wins from your A/B testing is to have a top notch system in place.

At Kissmetrics, our A/B testing program isn’t only about optimizing our conversions. We’re optimizing the system of testing…that eventually leads to more conversions.

If you want to learn how to get major A/B test wins, join our free email course to learn:

  • The proven system of testing we’ve used to triple our conversion rate (most companies spend years trying to figure this out)
  • A foolproof checklist for ensuring each test can lead to a big win (Yes, this can work for you if you apply it correctly. We’ll show you how)
  • What we learned from over 70 tests we’ve run on our website–including the #1 mistake even the most seasoned professionals make that could negatively impact your funnel

About the Author: Allison Carpio is the Product Marketing Manager for Kissmetrics.

$(function() {

Original Source File

CLIMBER B.C. Spring Summer 2017 | CPM Moscow by Fashion Channel

Source Article

The Funniest Wildlife Photos of 2016

The middle of October during a presidential election year is a really good time to remember not to take the world so seriously.

Case in point: The organizers of the Comedy Wildlife Photography Awards announced the finalists for the funniest animal photograph of 2016. The competition, in only its second year, is the antithesis of your traditional, staid wildlife photography contests. It’s a light-hearted competition that showcases the inner comedians in the animal kingdom, and it doesn’t t

Original Source File

How to Run an A/B Test in Google Analytics

Designs don’t always work out as intended.

The layout looks good. The color choices seem great. And the CTA balances clever and clear.


It’s not working. All of it. Some of it. You’re not completely sure, but something’s gotta give.

Despite everyone’s best intentions, including all the hours of research and analyses, things don’t always work out as planned.

That’s where continuous testing comes in. Not a one-and-done or hail & pray attempt.

Even better, is that your testing efforts don’t need to be complex and time consuming.

Here’s how to set-up split test inside Google Analytics in just a few minutes.

What are Google Analytics Content Experiments?

Let’s say your eCommerce shop sells Pug Greeting Cards. (That’s a thing by the way.)

Obviously, these should sell themselves.

But let’s just suspend disbelief for a moment and hypothesize that sales are low because you’re having trouble getting people into these individual product pages in the first place.

Your homepage isn’t a destination; it’s a jumping off point.

Peeps come in, look around, and click somewhere else.

Many times that’s your Product/Service pages. Often it’s your About page.

Regardless, the goal is to get them down into a funnel or path as quickly as possible, (a) helping them find what they were looking for while also (b) getting them closer to triggering one of your conversion events.

The magic happens on a landing page, where these two things – a visitor’s interest and your marketing objective – intertwine and become one in a beautiful symphony.

So let’s test a few homepage variations to see which do the best job at directing new visitors into your best-selling products.

One has a video, the other doesn’t. One is short and sweet, the other long and detailed. One has a GIF, the other doesn’t.


New incoming traffic gets split across these page variations, allowing you to watch and compare the number of people completing your desired action until you can confidently declare a winner.

(It’s probably going to be the one featuring this video.)

Running simple and straightforward split test like this is landing page optimization 101, where you identify specific page variables that result in the best results for your audience and multiply them across your site.

Google Analytics comes with a basic content experiments feature that will allow you to compare different page variations, split traffic to them accordingly, and get email updated about how results are trending and whether you’re going to hit your defined objective or not.

But… they’re technically not a straightforward A/B test. Here’s why, and how that’s actually a good thing.

Why Content Experiments Can Be Better than Traditional A/B Tests

Your typical A/B test selects a very specific page element, like the headline, and changes only that one tiny variable in new page variations.

The interwebs are full of articles where switching up button color resulted in a 37,596% CTR increase* because people like green buttons instead of blue ones. Duh.

(*That’s a made up number.)

There’s a few problems with your classic A/B test though.

First up, tiny changes often regress back to the mean. So while you might see a few small fluctuations when you first begin running a test, small changes usually only equal small results.

changes-regress-to-the-mean-ab-test(Image Source)

The second problem is that most A/B tests fail.

And if that weren’t bad enough, the third issue is that you’re going to need a TON of volume (specifically, 1,000 monthly conversions to start with and a test of at least 250 conversions) to determine whether or not those changes actually worked or not.

Google Analytics Content Experiments use an A/B/N model instead. Which is like a step in between one-variable-only A/B tests and coordinated-multiple-variable multivariate tests.

(After typing that last sentence, I realized only hardcore CRO geeks are going to care about this distinction. However it’s still important to understand from a high level so you know what types of changes to make, try, or test).

You can create up to 10 different versions of a page, each with their own unique content or changes.

In other words, you can test bigger-picture stuff, like: “Does a positive or negative Pug value proposition result in more clicks?”

Generally these holistic changes can be more instructive, helping you figure out what messaging or page elements you can (and should) carry through to your other marketing materials like emails, social and more.

And the best part, is instead of requiring a sophisticated (read: time consuming) process to set up to make sure all of your variable changes are statistically significant, you can use Google Analytics Content Experiments to run faster, iterative changes and learn on-the-go.

Here’s how to get started.

How to Setup Google Analytics Experiments

Setting up Content Experiments only takes a few seconds.

You will, however, have to set-up at least one or two page variations prior to logging in. That topic’s beyond the scope here, so check out this and this to determine what you should be testing in the first place.

When you’ve got a few set-up and ready to go, login to Google Analytics and start here.

Step #1. Getting Started

Buried deep in the Behavior section of Google Analytics – you know, the one you ignore when toggling between Acquisition and Conversions – is the vague, yet innocuous sounding ‘Experiments’ label.

Chances are, you’ll see a blank screen when you click on it that resembles:


To create your first experiment, click the button that says Create Experiment on the top left of your window.

With me so far? Good.

Let’s see what creating one looks like.

Step #2. Choose an Experiment

Ok now the fun starts.

Name your experiment, whatever.

And look down at selecting the Objective. Here’s where you can set an identifiable outcome to track results against and determine a #winning variation.


You have three options here. You can:

  • Select an existing Goal (like opt-ins, purchases, etc.)
  • Select a Site Usage metric (like bounce rate)
  • Create a new objective or Goal (if you don’t have one set-up already, but want to run a conversion-based experiment)

The selection depends completely on why you’re running this test in the first place.

For example: most are surprised to find that their old blog posts often bring in the most traffic. The problem? Many times those old, outdated pages also have the highest bounce rates.

Navigate to: Behavior > Secondary Dimensions + Google/Organic > Top Pageviews > Bounce Rate.

Here’s an example:


(Here are a few other actionable Google Analytics reports to spot similarly low hanging fruit when you’re done setting up an experiment.)

Let’s select Bounce Rate as the Objective for now, so we can make page changes to the layout, or increasing the volume and quantity of high quality visuals to get people to stick around longer.

After selecting your Objective, you can click on Advanced Options to pull up more granular settings for this test.


By default, these advanced options are off, and Google will “adjust traffic dynamically based on variation performance”.

However if enabled, your experiment will simply split traffic evenly across all the page variations you add, run the experiment for two weeks and shoot for a 95% statistical confidence level.

Those are all good places to start in most cases, however you might want to change the duration depending on how much traffic you get (i.e. you can get away with shorter tests if this page will see a ton of traffic, or you might need to extend it longer than two weeks if there’s only a slow trickle).

So far so good!

Step #3. Configure Your Experiment

The next step is to simply add the URLs for all of the page variations you want to test.

Literally, just copy and paste:


You can also give them helpful names to remember. Or not. It will simply number the variants for you.

Step #4. Adding Script Code to your Page

Now everyone’s favorite part – editing your page’s code!

The good news, is the first thing you see under this section is a helpful toggle button to just email all this crap code over to your favorite technical person.

If you’d like to get your hands dirty however, read on.


First up, double check all of the pages you plan on testing to make sure that your default Google Analytics tracking code is installed. If you’re using a CMS, it should be, as it’s usually added site-wide initially.

Next, highlight and copy the code provided.

You’re going to need to look for the opening head tag in the Original variation (which should be located literally towards the top of your HTML document. Search for to make it easy:


Once that’s done, click Next Step back in Google Analytics to have them verify if everything looks A-OK.

Not sure if you did it right? Don’t worry – they’ll tell you.

For example, the first time I tried installing the code for this demo I accidentally placed it underneath the regular Google Analytics tracking code (which they so helpfully and clearly pointed out).


After double checking your work and fixing, you should see this:


And now you’re ready to go!

See, that wasn’t so bad now was it?!


Websites are never truly done and finished.

They need iteration; including constant analysis, new ideas, and changes to constantly increase results.

Many times, that means analyzing and test entire pages based on BIG (not small) changes like value propositions or layouts. These are the things that will deliver similarly big results.

Landing page optimization and split testing techniques can get extremely confident and require special tools that only CRO professionals can navigate.

However Google Analytics includes their own simple split testing option in Content Experiments.

Assuming you already have the new page variations created and you’re comfortable editing your site’s code, they literally only take a few seconds to get up-and-running.

And they can enable anyone in your organization to go from research to action by the end of the day.

About the Author: Brad Smith is a founding partner at Codeless Interactive, a digital agency specializing in creating personalized customer experiences. Brad’s blog also features more marketing thoughts, opinions and the occasional insight.

Original Source File

YVES SAINT LAURENT SS 2000 Paris Fashion Week 4 of 5 pret a porter woman by FashionChannel

Source Article

Skin Cell to Egg Cell to Mouse Pup: Researchers Hail Fertility Breakthrough

Japanese researchers announced Monday that they used stem cells to create viable mouse egg cells entirely in the lab.

Scientists from the University of Kyoto coaxed skin cells into egg cells, which they then fertilized and implanted into female mice to successfully breed a new generation of mice. The technique had a low success rate, but it’s the first time lab-grown egg cells have been used to produce healthy offspring.

Previously, members of the same team had created mature egg cells

Original Source File