Data-Driven PR Measurement (eMetrics Chicago)

DATA-­‐DRIVEN  PR  MEASUREMENT  

Sandra  Fathi  President  Affect  @sandrafathi    

web:  affect.com  blog:  techaffect.com  email:  [email protected]    

eMetrics  Summit  Chicago,  June  9-­‐10,  2015  

Slides:  www.slideshare.net/sfathi  

ABOUT  ME  

•  Sandra  Fathi  •  President,  Affect  •  Public  RelaJons,  Social  Media,  

MarkeJng  

•  Council  of  PR  Firms  •  PRSA  Past  PosiJons:  

–  Tri-­‐State  Chair  –  NY  Chapter  President  –  Technology  SecJon  Chair  

•  WOMMA  

2  

 Technology Healthcare

Professional Services:    

SAMPLE  PAST  &  PRESENT    CLIENTS  

MEASUREMENT  &  METRICS  

Measurement Objectives 1.  Proving value of public relations activities 2.  Proving ongoing improvement in performance 3.  Securing headcount/budget for programs 4.  Demonstrating ROI compared with true business metrics

MEASUREMENT  &  METRICS  

Sample  PR  Key  Performance  Indicators  (KPIs):   1. Scores: Indices/scoring mechanisms to track valuable outcomes/results

•  Quantity: sheer volume of media hits •  Quality: score for Tier 1,2,3, score for feature, prominent, mention

2. Correlations: Between outputs, outcomes and business results. •  Track events with lead generation (online, email, phone, events) •  Track PR/social events with Web traffic

3. Check Boxes: Meeting specific, finite objectives

•  # of articles/month •  # of articles in target industries/vertical markets •  # of press releases per year •  # of members/attendees/downloads/registrations (hard numbers)

PR  MEASUREMENT  

Three  Concepts  for  Discussions:    

•  Share  of  Voice  •  CompeJJve  Benchmarking  •  CorrelaJons  

@sandrafathi   6  

PART  I:  SHARE  OF  VOICE  

DEFINITION  

Share  of  Voice:    

Comparing  your  crucial  performance  metrics  against  those  of  compeJtors  or  the  market.      •  You  have  to  measure  something  •  What  you  measure  needs  to  be  analyzed  proporJonately  against  compeJtor  data  (or  market  data)  to  establish  market  share  

@sandrafathi   8  

THE  FORMULA  

   Number  of  ConversaJons    

Including  Your  Company                              =    X  *  100  =  %  SOV  Total  ConversaJons  on  a  Topic  

@sandrafathi   9  

ADVERTISING  CONCEPT  

25%  SOV  

75%  SOV  

@sandrafathi   10  

SHARE  OF  VOICE  I  

72%  

28  %  

ConversaJons  

Talk  

About  Me  

@sandrafathi   11  

SHARE  OF  VOICE  II  

0%  10%  20%  30%  40%  50%  60%  70%  80%  90%  

100%  

Q1   Q2   Q3   Q4  

CompeJtor  C  

CompeJtor  B  

CompeJtor  A  

Our  Company  

@sandrafathi   12  

SHARE  OF  VOICE  III  

300  ArJcles  MenJon  My  Company  

145  ArJcles  MenJon  

CompeJtor  

589  Industry    ArJcles  

87  ArJcles  MenJon  Both  

51%  SOV    in  the  Industry  

@sandrafathi   13  

KEEP  IN  MIND  

•  Share  of  voice  should  be  defined  for  a  period  of  Jme  (finite  start  and  end).  

•  Share  of  voice  is  ojen  most  useful  when  limited  to  a  single  plakorm  or  medium.  For  example,  business  press  coverage  or  Twimer.  

•  Share  of  voice  can  be  overwhelming  if  trying  to  look  at  too  large  a  segment  or  industry.  Try  choosing  SOV  among  top  compeJtors  or  in  key  interest  areas.    

@sandrafathi   14  

SOV:  SOCIAL  MEDIA  ANALYTICS  PLATFORMS  

@sandrafathi   15  

SOCIAL  MENTION  

@sandrafathi   16  

SIMPLE  EXCEL  FORMULA  

@sandrafathi   17  

ONLY  PART  OF  THE  STORY  

•  Doesn’t  consider  senJment  

•  Doesn’t  consider  sources  (exclude  self  produced/owned  media)  

•  Doesn’t  consider  quality,  only  quanJty  (Is  NYT  blog  same  as  obscure  geek’s  tweet?)  

•  Don’t  accept  the  data  blindly  –  human  verificaJon  is  required  with  any  tool  

@sandrafathi   18  

OTHER  APPLICATIONS  &  CONSIDERATIONS  

ConsideraJons:  •  Apply  senJment  or  tonal  filters  (posiJve/negaJve)  •  Apply  qualitaJve  measures  (by  Jer  or  by  type)  

ApplicaJons:  •  Industry  trends/hot  topics  (i.e.  SOV  on  cloud  security)  •  Specific  products  or  services  •  Broken  down  by  geographic  or  demographic  parameters  

(i.e.  SOV  in  18-­‐25  market)  

@sandrafathi   19  

PART  II:  COMPETITIVE  BENCHMARKING  

DEFINITION  

CompeJJve  Benchmarking:    

The  conJnuous  pracJce  of  comparing  a  company’s  pracJces  and  performance  metrics  against  the  most  successful  compeJtors  in  the  industry.    •  You  measure  processes  and  results  •  You  must  idenJfy  a  ‘benchmark’  or  indicator  that  will  be  

a  unit  of  measure  to  compare  •  The  desired  outcome  is  to  understand  which  processes  

lead  to  greater  success  (best  pracJces)  in  order  to  improve  your  company’s  performance  

@sandrafathi   21  

COMPETITIVE    BENCHMARKING  

•  IdenJfy  my  compeJJve  set  for  comparison  •  Choose  my  units  of  measure:  press  coverage  •  Set  parameters:  top  20  business  and  trade  •  Define  a  Jme  period:  6  months  •  Choose  a  tool  (news  monitoring  service)  or  begin  manual  research  

@sandrafathi   22  

EXAMPLE:  RADWARE  

ObjecJve:      •  Build  &  Maintain  Radware’s  PosiJon  as  a  Thought  Leader  on  ADC  and  Security  

•  Maximize  Radware’s  Overall  Public  RelaJons  Results    Strategy:    •  Compare  and  Contrast  Radware’s  Press  Release  Output  with  Top  3  ADC  and  Security  CompeJtors  

•  Compare  and  Contrast  Radware’s  Coverage  with  Top  3  ADC  and  Security  CompeJtors  

•  Analyze  Results  •  Apply  Best  PracJces  and  Lessons  Learned  to  Radware  to  Improve  Overall  Performance  

@sandrafathi   23  

EXAMPLE:  RADWARE  

ApplicaJon  Delivery   Network  Security  

@sandrafathi   24  

•  Analysis  of  press  release  strategy  and  resulJng  coverage  over  6  month  period  

•  Specifically  as  it  relates  to  relevant  products  or  business  units  

•  Only  in  top  20  business  and  industry/sector  publicaJons  

METHODOLOGY  

@sandrafathi   25  

METHODOLOGY  II  

@sandrafathi   26  

RADWARE  PRESS  RELEASES  

Security  43%  

ADC  27%  

Both*  12%  

Other*  18%  

Press  Releases  

*  ‘Both’  includes  releases  related  to  both  security  and  ADC,  ‘Other’  includes  non-­‐product  releases  (e.g.  company  news,  financial  announcements  etc.)  

Press  Releases  

Security   14  

ADC   9  

Both   2    

Other   6  

@sandrafathi   27  

ADC  COMPETITORS:  SECURITY  &  ADC  

*  ‘Both’  includes  releases  related  to  both  Security  and  ADC  

11   11  2  

15  

46  39  

31  

92  

0  10  20  30  40  50  60  70  80  90  

100  

Radware   A10   Citrix   F5  

Press  Releases  

ArJcles  

PRESS  RELEASES  VS.  NUMBER  OF  ARTICLES  

@sandrafathi   28  

9   11  

2  7  

44  

16  

31  

38  

0  5  10  15  20  25  30  35  40  45  50  

Radware   A10   Citrix   F5  

Press  Releases  

ArJcles  

ADC  COMPETITORS:    ADC  ONLY  

PRESS  RELEASES  VS.  NUMBER  OF  ARTICLES  

@sandrafathi   29  

ADC  COVERAGE  BY  TYPE  

0  

20  

40  

60  

80  

Radware   A10   Citrix   F5  

Other  

Report  

Commentary  

AcquisiJon  

Partner  

Customer  

Product  

Product   Customer   Partner   AcquisiJon   Commentary   Report   Other  

Radware     7     2     1     0   1     34     1    

A10     1     1     10     0   1     0   26    

Citrix     27     0   0     0   2     0   2    

F5     24     0     3   0     12     0     37    

@sandrafathi   30  

COVERAGE  QUALITY  

0  

10  

20  

30  

40  

50  

60  

70  

80  

90  

100  

Radware   A10   Citrix   F5  

74%  

10%  23%  

50%  

26%  

90%   77%  

50%  

MenJons  

Features  

FEATURE  VS.  MENTION  

@sandrafathi   31  

ADC  CONCULSIONS  

•  Number  of  press  releases  did  not  correlate  to  number  of  arJcles  •  Radware  was  leading  in  SOV  on  key  topic  (ADC)  amongst  

compeJtors  and  the  quality  of  coverage  by  comparison  was  significant  (ValidaJon!)  

•  Overwhelming  majority  of  Radware’s  ADC  coverage  was  generated  by  reports  (ValidaJon!)  with  product  and  customer  news  trailing  far  behind  

•  CompeJtors  were  leading  with  product  news  and  capturing  media  amenJon  (Opportunity!)  

•  No  one  was  successfully  telling  the  customer  story  (Opportunity!)  

@sandrafathi   32  

SECURITY  COMPETITORS  

14  23   28  

10  

84  

164  

68   68  

0  

20  

40  

60  

80  

100  

120  

140  

160  

180  

Radware   Arbor   Imperva   Prolexic  

Press  Releases  

ArJcles  

PRESS  RELEASES  VS.  NUMBER  OF  ARTICLES  

@sandrafathi   33  

SECURITY  COVERAGE  BY  TYPE  

0  

50  

100  

150  

200  

Radware   Arbor   Imperva   Prolexic  

Other  

Report  

Commentary  

Amack  

AcquisiJon  

Partner  

Customer  

Product   Customer   Partner   AcquisiJon   Amack   Commentary   Report   Other  

Radware     8     9     6   0   28     22     11   2    

Arbor     19     1   1   18     29     44     52   0  

Imperva     2     1     1     0     8     18     19   19    

Prolexic     0     0   0   0     43     2     14     9    

@sandrafathi   34  

COVERAGE  BY  QUALITY  

0  

20  

40  

60  

80  

100  

120  

140  

160  

180  

Radware   Arbor   Imperva   Prolexic  

MenJons  

Features  

35%  

65%  

31%  

69%  

34%  

66%  

54%  

46%  

FEATURE  VS.  MENTION  

@sandrafathi   35  

SECURITY  CONCLUSIONS  

•  Radware  is  #2  in  overall  SOV  but  the  quality  is  not  as  strong  (more  menJons  vs.  features)  

•  Leading  customer  and  partner  conversaJons  (ValidaJon)  

•  Good  job  at  Story  Hijacking  (responding  to  security  hacks)  but  room  for  improvement  (ValidaJon)  

•  CompeJtors  winning  at  report  coverage  and  commentary  (Opportunity!)  

@sandrafathi   36  

CONSIDERATIONS  

•  Good  for  understanding  what  worked  but  not  necessarily  ‘how’  it  worked  

•  Costs  for  research  may  outweigh  benefits  of  insights  •  Once  you’ve  idenJfied  the  ‘best  pracJces’  you  may  or  may  not  be  able  to  replicate  them  

•  Consider  non-­‐compeJtor  companies  to  benchmark  •  Do  you  want  to  ‘emulate’  or  ‘innovate’?  

@sandrafathi   37  

PART  III:  CORRELATIONS  

DEFINITION  

CorrelaJon:    

A  mutual  relaJonship,  or  interdependence,  between  two  or  more  things.      •  In  the  absence  of  being  able  to  prove  ‘causality’  you  may  

be  able  to  demonstrate  a  ‘correlaJon’  to  demonstrate  the  impact  of  a  PR  or  markeJng  program  

•  A  correlaJon  is  posiJve  when  the  values  increase  together  

•  A  correlaJon  is  negaJve  when  the  values  decrease  together  

@sandrafathi   39  

TYPES  OF  CORRELATION  

Source:  MathisFun.com  

40  

THE  FORMULA  

41  

Pearson’s  CorrelaJon:  

@sandrafathi  

FUNCTION  IN  EXCEL  

42  @sandrafathi  

CORRELATION  IN  EXCEL  

43  @sandrafathi  

FUNCTION  IN  EXCEL  

44  @sandrafathi  

SCATTER  CHART  

45  @sandrafathi  

LINE  CHART  

46  @sandrafathi  

AcquisiJon  

3-­‐D  LINE  CHART  

47  @sandrafathi  

MULTIPLE  DATA  SETS  

48  @sandrafathi  

0  

500  

1000  

1500  

2000  

2500  

3000  

3500  

4000  

4500  

Q1   Q2   Q3   Q4  

Sales  

Web  Traffic  

Press  Coverage  

SPURIOUS  CORRELATION  

49  @sandrafathi   Source:  TylerVigen.com  

50  @sandrafathi   Source:  TylerVigen.com  

SPURIOUS  CORRELATION  

CONSIDERATIONS  

•  User  correlaJons  cauJously  and  don’t  trust  the  math  blindly  

•  The  visuals  ojen  tell  a  story  as  well  •  Remember  that  correlaJon  is  not  causality,  it  can  only  help  as  an  indicator  or  potenJally  predict  probability  

•  Data  is  sJll  bemer  that  your  opinion  

51  @sandrafathi  

FINAL  THOUGHTS  

•  In  measurement,  speak  the  language  of  the  C-­‐Suite  •  Excel  is  sJll  the  best  dashboard  for  data  visualizaJon  •  Don’t  be  afraid  to  learn  that  you  are  wrong  •  Don’t  be  afraid  to  change  direcJon  •  Use  the  data  to  gain  execuJve  support    

–  Strategy  –  Resources  –  Headcount  –  Budget  

52  @sandrafathi  

THANK  YOU    

CONTACT:  

Sandra  Fathi  President  Affect  @sandrafathi    

web:  affect.com  blog:  techaffect.com  email:  [email protected]    

Slides:  www.slideshare.net/sfathi