Attribution model choice in Adwords

Attribution is about putting the performance of your keywords into proper perspective.

Advertising should always target higher volume of profitable conversions. Choosing the right attribution model can help in improving advertising returns

Attribution across every single customer engagement is currently impossible.

However, choosing the right model can help in identifying successful keywords and messaging that play important role in generating business. This can help all parts of your marketing efforts, even beyond search and online advertising.

adwords attribution models compared

You can read more about selecting the most appropriate attribution model in this Google support document  https://support.google.com/adwords/answer/7002714

Voice search, Artificial intelligence, PPC and SEO

How voice search might change PPC

The voice search volume is growing many folds. Many would agree that voice is a more convenient way to interact with computers- its more natural than typing text or using a mouse.

Today when you search by voice you get search results on your mobile or desktop device same as if you might have typed the search query.

The input is voice but the output is still text based. The online text and image ads are still showing.

However soon users might prefer the output itself to be in voice format.

Voice output could be more convenient and preferable because it would allow users to multitask while searching, thus saving time. Example a user might search by voice while driving or exercising or having his or her meal. When the output would be voice, the online ads as we know them today, will not be “visible” anymore.

Echo Dot was the best seller in November 2016.

Devices like these and the rapid improvements in the voice interface of mobile and desktop operating systems will certainly change the search and PPC world.

Purna Virji, Senior Manager, PPC Training at Microsoft wrote an excellent article about “How Voice Search Will Change Digital Marketing“. This article is very useful to understand how voice search will change the input side of search as we know it today. However it does not talk much about the output side – that is, how we consume search results.

Voice search will make top results even more important

Today, we are used to seeing the search results on a screen, scanning them fast and choosing between them.

In my opinion and as I see in my client accounts, the top positions of search result pages always enjoy significantly higher traffic and business than search results at the bottom of the search result page.

Page 2 of search results gets even less traffic than page 1 and that too would be from people who want to research their decision a lot, otherwise in most cases the top results on page 1 of search engines suffice for most users, especially for most commercial queries.

When voice search becomes popular and the user would start expecting the output to be in voice format too, then the top result that the “computing device” outputs would become even more important than the top result in the search result pages as we know them today.

Our eyes are far faster in scanning more results as compared to waiting for the computing device (the virtual assistant) to speak the next listing details.

So the “second listing” in search results might start getting even fewer chances.

Add to that the “artificial intelligence” technologies that are rapidly growing today. Our “digital assistants”will know more and more about our context and preferences, so the results are bound to get more and more personalized.

The consumption of search results via voice and the artificial intelligence algorithms constantly trying to learn each individual’s preferences will definitely make huge impact to the PPC and SEO landscape as we know it right now.

If today the web marketing world is 70/30 or 80/20 – the top 2 or 3 listings get majority of the business and clicks, then in the voice world this might be 90/10 – the top listing might start getting far more business share than the “listings” below it.

I cannot be sure and I could be wrong but I can imagine that users may not like to wait to “listen” to the second option, especially if the “assistant” keeps learning and in most cases presents very relevant results in the first “listing” itself.

It would become like a habit to accept the top listing – “the answer” that the assistant is providing.

On radio, most users are not really expecting any “results” to their query. Radio is mostly just entertainment and ads in between the content are digestible for most users. How many pay attention to such ads on radio or even on TV is a matter of growing debate, especially when we see around us that as soon as ads come, many users pick up their mobile or start doing something else.

However, imagine a scenario where you are communicating with your voice-assistant, seeking answers to your query. You would want the answer right now. You don’t want to hear three ads before your result (especially if you think the ads may not be relevant). In text search, you can easily scan past ads in case if you feel the ads are not relevant. However voice output would make embedding ads a bit more tougher.

More personalized and context aware ads, at more suitable timings.

On the flip side however, the personal voice assistants would know a lot more about us, than today’s search engines. So I am sure there will be advertising opportunities due to the “predictive response” behavior which we will start expecting from our virtual assistants in the “internet of things” world.

The virtual assistants might be able to guess when it is time to order something, as well as, what sort of activity we are involved in at a given point of time and whether we are receptive to ads at a given point of time.

For example, the virtual assistant could guess that we are on a routine walk, and communicate with us if we want to know about the promotions available for the next purchase due soon. We might consider listening to these “promotional messages” at such a time. Otherwise the assistant can learn that we are not interested in listening to messages when we are walking and try some other time, till it understands which time is suitable for what type of message.

These type of interactions and the context aware nature of virtual assistants could make new advertising chances available at most suitable timings.

So far, voice search is very new to me. I find voice search and AI fascinating and will definitely be thinking, reading and sharing more thoughts about it in near future.