Format Negotiation with Buyers

[Benji Weber]

Hello Everyone,

We’re interested in being able to negotiate ad format capabilities with buyers. In some cases the spec can support this already. However, it’s inconsistent and often the only way to do it seems to be to send multiple impressions out - one for each permutation of options we support, which is wasteful and non-scalable.

For example we might support on a single page impression multiple:

• Formats (Width, Height, Ratio) e.g. we could accept a 16:9 creative or a 9:16 creative in the same position on the page.
  
• Playback Method e.g. we could support playing back when clicked or when 100% in view or 50% in view, depending on what the buyer wants.
  
• Playback Cessation
  
• Video Placement Type e.g. could be Floating or In-Article or Hero potentially.
  
• Payment event 
  
• ...
  

One can already do something similar with Skippability - advertise skippability and the bid can include creative attribute 16 to request it.

I wanted to check 

1. I’m not missing an obvious way to do this kind of negotiation already (other than sending multiple imps
   
2. Is this something that would be useful ?
   
3. Is anyone working on a proposal for improving capabilities in this area already?
   

I'm interested in working on this or working with someone on this if not.

Thanks

--

benji

The contents of this email are confidential and the information in it may not be disclosed except for the purpose for which it has been sent. If you have reason to believe that you are not the intended recipient of this communication, please contact the sender immediately.

Unruly Group Limited - Registered in England No. 5411297, VAT #: 108 2572 28
Registered Office: 42-46 Princelet Street, London, E1 5LP, UK

[Ian Trider]

Hi Benji,

1. You're not. At least for my part in this (i.e. placement type, skippability) this was intentional because I believed negotiation to be an obscure use case that rarely applied.  I failed to anticipate the flexibility of outstream, for example. 

2. Possibly, though I fear it will be painful to implement (adds complexity). 

3. Its on my wishlist for OpenRTB 3.0 (as this would almost certainly have to be backwards-incompatible), but I haven't written anything specifically yet.  Perhaps we could collaborate on this.

[Benji Weber]

Hi Ian,

Good to know. Yes it would be good to collaborate on this.

I'll think through some of our needs and potential approaches in more detail and then let's discuss.

--

benji

--

benji

--
You received this message because you are subscribed to the Google Groups "openrtb-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openrtb-dev+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Benji Weber]

Hello Everyone,

I've been discussing this a bit more internally and have mostly come up with more reasons why it's difficult rather than any approach I'm particularly keen on.

I've started a google doc with things we've thought about so far in case anyone has any bright ideas or would like to comment https://docs.google.com/document/d/15_zbx1lQYA-cQ38TyBr0v2ptu4nh7jp_lHP6rDt9I1M/edit?usp=sharing

Thanks

--

benji