5 ms·

As an engineer, I have always wanted to quantify the effect of PR. The PR people I spoke to often says it is hard to measure, occupying mindshare, "upper" funne

by throwaway_2047 5y ago

As an engineer, I have always wanted to quantify the effect of PR. The PR people I spoke to often says it is hard to measure, occupying mindshare, "upper" funnel, yada yada. Some may offer number of publicity as a measuring stick. Would love to learn what other "more useful" metrics are out there.

ghaff 5y ago
One of the challenges is that the most obvious numerical measures are mostly relevant to crappy PR, i.e. the spray and pray types that often seem to fill my inbox. The better PR people I know are much more about targeted relationships and their output metrics are probably something more along the lines of coverage in "good" (a word doing a lot of work here) outlets.
- stadium 5y ago
  I believe that those emails would be considered marketing, not PR. https://www.themuse.com/amp/advice/the-difference-between-marketing-and-pr https://www.themuse.com/amp/advice/the-difference-between-ma...
  - ghaff 5y ago
    It's a mix. Yes, I get a ton of "nurture" emails for webinars and the like. I also get a ton of PR pitches that want me to write about something or interview some "expert" about a topic. (I have a blog, a podcast, and write or have written for various online pubs so I'm on a ton of mailing lists.)
mkmk 5y ago
You might be interested in Media Mix Modeling — a regression-based approach to finding relationships between marketing spend and sales performance. A good primer is here: https://github.com/mecommerce/ThirdLove-Tech-Blog/blob/master/Media_Mix_Model/ThirdLove_MMM_Whitepaper.pdf https://github.com/mecommerce/ThirdLove-Tech-Blog/blob/maste... And https://www.vexpower.com/ https://www.vexpower.com/ is an early but very promising educational tool in the space.
- hammock 5y ago
  Media mix modeling is well-intentioned but in reality a huge overpromise and highly profitable way for data scientists to snow/fleece their advertiser clients. (More detail.. it's a bit like the Fed's inflation data: when you can manipulate the constants at will, introduce fake data as proxy for missing data, etc.. the ways and means to force a desired result are endless)
  - mkmk 5y ago
    There’s a lot of truth to that. When done in house, though, it can be a useful additional check on spend allocation. Of course it still becomes political if channel owners feel like their career progression depends on growing share of budget.
AlbertCory 5y ago
Read a classic: "Ogilvie on Advertising" (1985). The only metric that matters is: did your sales go up? Ogilvie understood that. The ads that other PR and advertising people admire aren't necessarily the ones that work.
- nickelcitymario 5y ago
  Pedantic correction: "Ogilvy". While the only metric that matters is sales (I fully agree), you need other metrics to understand how those sales happened. This is the dangerous thing. I've worked at companies where people were evaluated based on very specific metrics, like web traffic or social media engagement. The problem with that is you end up focusing on improving those numbers rather than improving sales. But if you only focus on sales, you can't tell which activities made a difference (or in which direction). You need other data to understand what's happening. A better book than Ogilvy on Advertising, in this regards, is Claude C. Hopkins' Scientific Advertising. It lays out all the principles for testing ads, back when we didn't have any concept of analytics. The basic method was: Try different things in different markets to see what works best. Keep doing it indefinitely, so that your ads only ever get better. There's no reason the same approach can't be taken with public relations. If you're working for a national brand, try assigning some PR folks to exclusively work the media in one given region. Then see what the difference is. (Obviously, you have to do a lot of work to "control the variables", but any data scientist -- or any scientist at all, actually -- should be able to figure that out in their sleep.) It's not rocket science. These methods have been around for a very long time. Honestly, I think we've just gotten lazy. Setting up market tests is a lot of work, whereas creating an A/B test in AdWords is easy. But there's no such thing as an unmeasurable publicity tactic (whether it's advertising or public relations). You just gotta be willing to do the work.
  - sosborn 5y ago
    With PR, the effects of a failed test can take down a company (or, the existing c-suite at least). That's why people tend to stick to a "tried and true" playbook.
    - 5y ago

ghaff 5y ago

One of the challenges is that the most obvious numerical measures are mostly relevant to crappy PR, i.e. the spray and pray types that often seem to fill my inbox. The better PR people I know are much more about targeted relationships and their output metrics are probably something more along the lines of coverage in "good" (a word doing a lot of work here) outlets.

stadium 5y ago
I believe that those emails would be considered marketing, not PR. https://www.themuse.com/amp/advice/the-difference-between-marketing-and-pr https://www.themuse.com/amp/advice/the-difference-between-ma...
- ghaff 5y ago
  It's a mix. Yes, I get a ton of "nurture" emails for webinars and the like. I also get a ton of PR pitches that want me to write about something or interview some "expert" about a topic. (I have a blog, a podcast, and write or have written for various online pubs so I'm on a ton of mailing lists.)

mkmk 5y ago

You might be interested in Media Mix Modeling — a regression-based approach to finding relationships between marketing spend and sales performance. A good primer is here: https://github.com/mecommerce/ThirdLove-Tech-Blog/blob/master/Media_Mix_Model/ThirdLove_MMM_Whitepaper.pdf https://github.com/mecommerce/ThirdLove-Tech-Blog/blob/maste... And https://www.vexpower.com/ https://www.vexpower.com/ is an early but very promising educational tool in the space.

hammock 5y ago
Media mix modeling is well-intentioned but in reality a huge overpromise and highly profitable way for data scientists to snow/fleece their advertiser clients. (More detail.. it's a bit like the Fed's inflation data: when you can manipulate the constants at will, introduce fake data as proxy for missing data, etc.. the ways and means to force a desired result are endless)
- mkmk 5y ago
  There’s a lot of truth to that. When done in house, though, it can be a useful additional check on spend allocation. Of course it still becomes political if channel owners feel like their career progression depends on growing share of budget.

AlbertCory 5y ago

Read a classic: "Ogilvie on Advertising" (1985). The only metric that matters is: did your sales go up? Ogilvie understood that. The ads that other PR and advertising people admire aren't necessarily the ones that work.

nickelcitymario 5y ago
Pedantic correction: "Ogilvy". While the only metric that matters is sales (I fully agree), you need other metrics to understand how those sales happened. This is the dangerous thing. I've worked at companies where people were evaluated based on very specific metrics, like web traffic or social media engagement. The problem with that is you end up focusing on improving those numbers rather than improving sales. But if you only focus on sales, you can't tell which activities made a difference (or in which direction). You need other data to understand what's happening. A better book than Ogilvy on Advertising, in this regards, is Claude C. Hopkins' Scientific Advertising. It lays out all the principles for testing ads, back when we didn't have any concept of analytics. The basic method was: Try different things in different markets to see what works best. Keep doing it indefinitely, so that your ads only ever get better. There's no reason the same approach can't be taken with public relations. If you're working for a national brand, try assigning some PR folks to exclusively work the media in one given region. Then see what the difference is. (Obviously, you have to do a lot of work to "control the variables", but any data scientist -- or any scientist at all, actually -- should be able to figure that out in their sleep.) It's not rocket science. These methods have been around for a very long time. Honestly, I think we've just gotten lazy. Setting up market tests is a lot of work, whereas creating an A/B test in AdWords is easy. But there's no such thing as an unmeasurable publicity tactic (whether it's advertising or public relations). You just gotta be willing to do the work.
- sosborn 5y ago
  With PR, the effects of a failed test can take down a company (or, the existing c-suite at least). That's why people tend to stick to a "tried and true" playbook.
  - 5y ago