2:50 — Kurt Vanderah introduces Ant’s Eye View’s Kira Wampler.
2:51 — Was at Intuit for 6 years, until 3 months ago. Accountants not only buy from us, they’re responsible for recommending 40% of our business. I became responsible for figuring out how to measure this, and became our first WOM person, in charge of community, etc.
2:52 — Measurement is a journey. We need to set expectations when we start that we’re early and experimenting.
2:53 — If you don’t set a baseline, you won’t know how your efforts are going. If you do that, your efforts are scaled, predictable and resourced.
2:54 — You may not be scaling or driving results or you be too strict – matrix in her first slide.
2:55 — Channel health measures are critical. But imagine if you went to senior management and said “10 million people saw our ads.” And then you stopped. This doesn’t connect to business metrics. Senior leaders want to see how you connect these to your business impact. They want to know how to think about this vs. the other decisions they make as a leader.
2: 57 — There are 4 approaches to making a measurement connection –
2:57 — Table Stakes:
1. Behavioral
2. Claimed
2:58 — Sophisticated Approaches:
1. Testable
2. Datamining
2:59 – Behavioral: I’ll believe it when I see it.
2:59 — How many of you have put your Social media urls into your website analytics? Small percentage of the room.
3:00 — If you’re not putting your urls into your analytics suite then you haven’t started yet. If you are working with a large company with a global suite vs business unit suite, some units will measure differently. Example: Anything “unclassified” was counted as WOM on Team A while with team B, only Facebook traffic was WOM
3:01 — Product adoption can be tracked via Social Marketing. An example is Dell Outlet Twitter codes. When page level analytics aren’t available, such as Amazon reviews, we don’t know the funnel. When Social Media is “Part of the process” – tough to measure. Sometimes channels are not big enough to drive statistically significant results.
3:02 — Example Behavioral campaign dashboard slide: Always tie to business results.
3:03 — Claimed: The opposite of behavioral, or I’ll believe it when the survey says it.
3:04 — Ask how social media or online reviews influence the purchase in a survey. Don’t think of social media or WOM, that is another entity. Many of us survey our customers all the time. Use that data to better understand things like the impact of social media on the purchase process.
3:05 — Amazon is a huge impact on sales. We had to have 100% reply rate on Amazon reviews. We had first product creator inline responses on Amazon. Quickbooks Pro 2010. Important because we showed that we cared about customers and answered questions immediately.
3:06 — A problem occurs when it is the right thing to do but the impact is unknown. We used data from 3 different customer surveys to triangulate impact of the reviews.
3:07 — We learned that the impact was 3-12% of sales impacted by reviews. This is a multi-million dollar impact. Our senior managers were able to use this in financial models.
3:08 — Testable: I’ll believe it when it is significant.
3:09 — A/B testing websites with engagement and functionality. When we included online engagement we saw increase in revenue. Showing a link to community would increase revenue vs. the page that didn’t have it.
3:10 — Message-test twitter message for reach, click through and conversion.
3:11 – Datamining: I’ll believe it when I regress it.
3:12 — We did matching community profile to customer data and saw that 30% of new community members bought an Intuit product within 24 hours of joining community.
3:13 — 2 things you can do tomorrow:
- Classify all relevant Social URLs in your analytics tool
- Ask customers about their purchase process and what influences it
3:14 — That will help you move social activity into the business process.
Q & A
Q: Laurie Marino from Thompson/Reuters, “Looking at Omniture I can see Twitter and Facebook traffic, but I can’t put my javascript code on those pages.”
A: Use Claimed when you can’t get page level analytics. Ask about the impact of social media on customer purchase. Use custom links. Use Coupon code. But this still doesn’t help with Universe or Funnel size. Use % of how many people online via Claimed data, and % use twitter, etc. Hoping that these guys are working on deeper analytics tools.
Q: “A lot of clients are smaller software companies – they’re trying to replace pay per click with social media.”
A: Go back to behavioral vs claimed. We can see how many people click on a link and the conversion rates. To be frank, the senior leaders only cared about the actual behavior – related to revenue. Often social isn’t the last step in a funnel – so I like both.
Q: Matthew from H&R Block, “How do you handle chicken and egg problem? Community members are more likely to adopt vs Adopters join communities.”
A: Testing is the best. What we saw in small business group, we were tracking engagement elements, to see how many people would engage. Versus control case without engagement elements, if we didn’t show engagement, we proved those people were less valuable from a revenue perspective.
Q: Trish from Mabels’ Labels, “We don’t have expertise to do regression analysis in house. What do you recommend?”
A: Google Analytics are used by many teams and small companies. This can help you with funnel management and conversion flow. Get interns in Grad Programs with analytical background. Working with a person on analytics – be clear about what questions you’re asking for from the data. Figure out what you’re trying to achieve and how can data help you figure out if you’re missing something.
Love this live coverage? It’s all thanks to the hard work of the very talented Howard Greenstein.
Procter & Gamble: Swan dive… into Old Spice’s remarkable social media campaign
Alberto Culver: Managing promotions on Facebook
UnitedHealth Group: Success in social recruiting
BlogWell: How Big Brands Use Social Media