Live blogged from the 2015 MRIA National Conference in Toronto. Any errors or bad jokes are my own.
Panel – Digital and Social Media impact on Research
Twitter/Google/Facebook/Rogers – Sponsored by Ipsos
Gayle Lunn, Ipsos UU
- Industrial revolution, mass production/consumption which changed the quality of people’s lives, digital now lets us to do things differently and impacts the way we think
- Luke Stringer from Twitter – startling change joining Twitter – speed of things changing is very fast, challenges the research method. Product is very different today from what it will be in one week or one month. Culture is very collaborative, work spaces are collaborative, completely open concept [booooooooooooooo!] There are tables around so people can meet and talk. Culture of failure is embraced, test and fail fast. They test every minute change to the product.
- Alexandra Cohn from Google – Came from ipsos, overwhelming amount of data and knowledge. Like drinking from a firehouse. Acronyms for everything. What do you need to know and what can you ignore. Speed is overwhelming. Completely different mentally, not asking, the data is behavioural. Fully transparent and collaborative culture, always available to anyone. No one keeps data to themselves.
- When you join twitter, you give your name and email. They don’t know who you are as an individual. Why do you use twitter – they don’t know that. You can apply traditional techniques to understand that. You have to ask questions.
- Not everything can be done in no time. There’s a lot of pressure to do research fast, get answers cheaply.
- Will google surveys replace traditional surveys?
- Collecting too much data is a bad things, especially from consumers.
- Despite all the internal knowledge, they still need to pair internal data collection with elicited data. It’s enrichment of data.
- Google surveys is only ten questions by design. Partnering with the full service companies is where they get their strength because google isn’t researchers. They provide the data, they are the platform. Still need researchers to interpret the data, researchers know how to ask the good questions. Not just anyone can ask a question.
- Twitter is a mobile first company. Rich media is limited, on purpose. There is always a place for longer form surveys, rightly or wrongly.
- Data scientist is a title that has to be earned.
- A lot of collaboration is behind the scenes, not public, and you can’t talk about it. But there is a lot of collaboration among clients. Collaboration is driven from the outside. Clients can’t operate in silos anymore, they need data from a wide range of people.
- Digital is held to a higher standard than other media. People expect you can link data but you can’t always do so. You can’t always use JUST data to complete a model. Big expectations that models will drive income.
- Right now we pay for view but maybe we will eventually pay only if someone follows or only if someone makes a purchase – pay for desired behaviour.
- Like to append their data to market data to see true ROI. Research explains why one campaign worked and another did not. Which specific type of content drove the result.
- Just “knowing” something works is not enough. They want and can get precise numbers.
- What is the value of a Twitter like versus a Facebook like? Need to be able to measure this. Models are wonderful but they don’t matter. Metrics matter.
- TV ads existed for a long time and we had ways of measuring the outcome, same with print and radio. Twitter has to figure it out for themselves instantly, not over decades.
- Industry is evolving but not fast enough. Privacy laws are very challenging. Internal struggles between product development and engineers, all have their own internal objectives.
- Rely on industry to tell them what the cool things are, where the industry is going. Partnered with academics to understand impact of being exposed to twitter versus other media in terms of how brains are reacting. Have and are exploring neuroscience to understand their product.
- Investing in machine learning and artificial intelligence. If you want to learn the direction of the company, look at who is on their board.
- How do i read irony in a google search query? 2/3 of top AI people in the world work for google.
- How can we collaborate more effectively. External partners come with ideas as opposed to we just want to work with you.
- It doesn’t have to be unique data. We just need to properly marry the data.
- It’s okay to fail in this safe space, with appropriate checks and balances
- Sometimes, retweets and shares are the right measure for you. Other times, ROI is the right measure.
- How do you attribute a click when someone saw six ads along the way and then finally clicked on one.
- Collaboration can lead to indecision because no ones in charge. Must hold everyone accountable for specific objectives. Make it directly related to reviews. Course correction is ok. Autonomy is important.
- Collaboration is not leading my consensus. One person must be responsible for the outcome.