Dumb question, but I'm not a big PPC guy and haven't dabbled in re-marketing before.

I work on a group of sites. There are a number of informational ones and a single more commercial one. They are all very publicly part of the same thing. Looking at things it looks a lot like I can retarget visitors to the info sites with a campaign promoting the money site.  Is that correct? It seems a bit open to... err "creative uses"

>Is that correct?

That is correct

Wow. Thanks.

Now to decide whether to ask permission or beg forgiveness! Does anyone really read privacy policies??

http://www.seroundtable.com/google-remarketing-wider-15217.html

Keep a light hand about remarketing. It can be very effective if your targeting is good but it can be easy to make the user feel stalked if you overdo it.

Suggestion: Set frequency limits  (X number of impressions per day/week/whatever) so the user sees your ads occasionally but isn't bombarded. Also, use category exclusions to weed out the most irrelevant exposure. An add for knitting supplies appearing on a site about truck parts would seem like a real misfit, while the same ad on a gossip site would be a lot less likely to contribute to creepy feelings.

The creepiness thing is a big factor for me. Would definitely keep in mind, particularly as the client I am thinking of has a whiter than white brand.

Theoretically you could then re-market to anyone who's viewed a facebook app/tab.  Very interesting. Anyone tried that?

Any success stories with G's re-marketing?  I mean, it sounds great, but has anybody made it convert?

I have some experience with remarketing, and the traffic converts well.

One campaign in particular was responsible for about 1% of the site's traffic but about 3% of the sales.

The campaign targeted people who had put something in their shopping cart but didn't complete the purchase. Those people responded well to remarketing, even though we weren't offering any kind of special promotion, just our ordinary banners.

We set tight limits on how often those users would see our remarketing campaigns, but we bid aggressively to make sure we did in fact get those impressions.

The limits were partly to avoid the creepiness factor, but also, your CTR stays stronger if you don't over-expose your ads and that helps to keep your costs in line.

In my experience that applies to all AdWords campaigns, not just remarketing.

In fact, being able to set frequency caps would be my biggest wish for buying ads on Facebook!

retargeting rocks 

I'm a little more cynical about retargeting, having tried Criteo, Struq and Google. Measured on last click, it worked. It looks even more impressive if you only show ads to those who've already shown signs of converting. How about measuring on first click? It doesn't bring in any new visitors to your website at all, so won't have any sales attributed to it. It's the marketing equivalent of someone who stands near the goal in a game of school football (soccer), taps in the ball and claims the glory; an advertising goal-hanger. That doesn't mean it isn't a valuable role, but it can't be measured in the same way acquisition channels are measured.

I think to really measure whether retargeting's worthwhile you need two things:

1. The ability to split test: Are those seeing the retargeting ads showing a higher propensity to convert? The retargeting companies will promote the fact that you're getting free publicity unless someone clicks an ad, but you need to remind yourself that the only people seeing these ads are those who already know about you. As such, it's an entirely different model that contextual display ads.

2. A distributed attribution model: Retargeting is potentially the last touchpoint in the purchasing funnel. If it works, it's about improving conversion, not about acquisition. As such it should be treated as an overhead, not an acquisition channel. The same could be said about email, and you'd expect a far better ROI from that than an acquisition channel like non-brand PPC.

we do run it with attribution, as far as we can. fundamentally we can see most touchpoints in a journey, but there's a lot of argument about the weighting to apply, especially when its a multi agency thing.

I think at the basic level there's no client so far where we haven't found conversions improve with reargeting, assumng static traffic. Criteo also works best in 95% of cases, so far, and personalised, highly relevant retargeting and products works far better than simply branded retargeting.

I may be about to much up the ROI slightly as I'm now experimenting with forcing some social traffic through the retargeting loop (not using the above, nice idea though ) and with less engaged visitors (ie; ones only there because they wanted something funny or to win a prize) we're clearly into the realms of lots less engagement and potential accusations of stalker.  I think it might be really hard to measure this accurately too. It's definitely worth a try though.

Quote from: jetboy on June 28, 2012, 03:36:14 PM
1. The ability to split test: Are those seeing the retargeting ads showing a higher propensity to convert? The retargeting companies will promote the fact that you're getting free publicity unless someone clicks an ad, but you need to remind yourself that the only people seeing these ads are those who already know about you. As such, it's an entirely different model that contextual display ads.

That's a very good point - if I've visited a site and don't want to buy from it, I'm followed around by the ads regardless.

If I've visited a site, and decided to buy from it later, the ads still follow me - even after purchase.

As a user it makes me personally cynical that remarketing is as useful as claimed. People claim they have stats to show it works - but if I'm buying I'm doing that anyway, regardless of remarketing. How does that factor into their positive assertions of it being a successful strategy when it facillitated nothing, but still claims the glory?

With jetboy's analogy, I'd suggest further - that in football it's like the person nearest the net being attributed the goal, even if they never touch the ball. Common sense says they will some of the time, but also that remarketing is wildly exaggerating its goal count.

its not always last click actually - although its normally reported as so because a lot of media agencies report on 'last click' but actually claim last click as anything which viewed an ad and still shows the ad cookie. I don't understand how they justify this, they just do it in a lot of cases.

What normally converts for higher value product is a brand search term. 

I don't especially like retargeting when I see it happening to me, I never click on banners, but don't mistake the fact it doesn't convert you for the fact it doesn't convert at all! I am a cynic who's seen the light on regargeting and it absolutely increases conversions if its done properly.

QuoteI'm followed around by the ads regardless

That "followed around" feeling is why it's so important to set impression caps, IMHO.

Quoteeven after purchase

The smart remarketer will suppress ads for people who have already bought something, or show them something different.

Learn how to create custom audiences. It's the secret to finesse.

Quotea lot of media agencies report on 'last click' but actually claim last click as anything which viewed an ad and still shows the ad cookie

I inherited an email agency that reported sales based on anyone who'd *received* an email. At that time we were sending blanket emails to everyone we had an email address for. As such, the majority of repeat bookings were attributed to email. They got quite aggressive when I suggested that they might be taking the piss. We increased legitimate conversions from email tenfold within the first year after binning them.

QuoteThe smart remarketer will suppress ads for people who have already bought something

Absolutely. Not many of them about. Present company excepted of course. I do think it'll work for us, but the tests taught me that we don't yet have the right tools in place to measure it intelligently.