Takeaway: Our concerns used to be about lofty expectations, now the story is getting hairy. How much longer will the street pay up for fake users?

NOTE SUMMARY

  1. CRUSHED 2Q14, BUT NEW CONCERNS: TWTR beat the midpoint of its 2Q14 guidance by a whopping $37M (13.5% upside).  Both revenue segments accelerated on a y/y basis, and US User growth sharply accelerated (see below).  However, there may be an emerging trend around ad engagements that may mean its runway is shorter than initially believed.
  2. GUIDANCE RAISE DRIVEN BY ACQUISITIONS?: TWTR spent $134 million on acquisitions in 2Q, all within its Data Licensing/Other Segment, which saw considerable surge in revenue in 2Q (up 43% q/q vs. 6% in 1Q).  TWTR raised revenue guiance by ~$100M; but 40M came from 2Q upside to guidance.  The question is how much of the raise is organically driven when the Data segment saw a $10M surge in 2Q revenue alongside intra-quarter acquisitions.  Waiting for the 10-Q on this one, stay tuned.  
  3. USER NUMBERS ARE EXAGGERATED: This is the bigger story.  TWTR has a large number automated accounts (Bots) that link into twitter.  These aren't real users, so can't be monetized.  More importantly, bots are growing as a percentage of its total, and have represented a substantial portion of its user growth over the past year (+40% of new MAUs), which begs the question, what is the real runway here? 

CRUSHED 2Q14, BUT NEW CONCERNS

TWTR beat the midpoint of its 2Q14 guidance by a whopping $37M (13.5% upside).  Both revenue segments accelerated on a y/y basis, with advertising revenue growing 129% y/y (vs. 125% in 1Q), and Data Licensing & Other up 91% (vs. 76% in 1Q).  The big surprise was the acceleration in US User growth, which inflected higher after 6 quarters of decelerating growth (however, there may be some distortion, see next section for more detail).  At face value, just a very solid quarter.  

Another positive is that new ad formats and stronger advertiser demand around the World Cup led to its first q/q increase in cost per ad engagement (CPE).  However, total ad engagements decelerated considerably q/q to 4% (vs. 28% last quarter).  At face value, the combination of these two factors could suggest that TWTR isn't stuffing the channel with rising ad load as aggressively as we expected. 

However, it could also mean that TWTR may still be stuffing the channel, but the engagement rate on that rising ad load is declining.  2Q14 marks the first quarter in its reported history where the sequential growth in timeline views outpaced that of ad engagements.  If the ad engagement rate is declining, it would suggest that TWTR users are starting to fade/ignore ads. In turn, rising ad engagement would increase at a proportionately lower rate than the increasing ad burden it placing on its members.  In short, the attrition risk rises for a waning yield. 

TWTR: Fake User Growth to Get Worse (2Q14) - TWTR  Timeline vs. Ad Engagements 

GUIDANCE RAISE DRIVEN BY ACQUISITIONS? 

TWTR spent $134 million on 7 acquisitions in 2Q, all within its Data Licensing/Other Segment, which saw considerable surge in revenue in 2Q (up 43% q/q).  TWTR raised guiance by ~$100M; but ~40M came from 2Q upside to guidance, which translates to roughly $60 increase to its 2H14 outlook.

The question is how much of the raise is organically driven when the Data segment saw a $10M q/q surge in 2Q revenues alongside 7 acquisitions that were added intra-quarter, hence didn't contribute a full quarter of revenues.  2 more were announced in this quarter, both of which are likely factored into guidance as well, and we could see more coming.  It's possible that we could see a meaningful shift in revenues toward its Data segment next quarter.  We're waiting for the 10-Q to settle this one, so stay tuned.  

USER NUMBERS ARE EXAGGERATED

This is the bigger story.  TWTR has a large amount automated accounts that link into twitter with no user interaction ("bots").  These bots aren't real users, they do not engage in content, so they can't be monetized (don't click on ads).  

During 2Q14, TWTR disclosed that these bots represented 14% of its MAUs (38M); that is double the percentage from 2Q13.  Further, these bots represented a substantial portion of its user growth over the past year.  Total MAUs increased by 53M y/y in 2Q14, 23M of those were bots (+43%).  After backing out bots from its MAUs in 2Q14, real user growth was up only 15% y/y, vs. the 24% reported by management.  On a q/q basis, TWTR's real user growth has dwarfed that of total user growth over the LTM, despite the relatively smaller user base

TWTR: Fake User Growth to Get Worse (2Q14) - TWTR   Real vs. Total User Growth

Management comments suggest this may be a growing issue.  Timeline views per MAU have been declining on a y/y basis over the past 3 quarters as bots represent a growing percentage of total MAUs, in turn, inflating the denominator. Management stated during the 2Q14 call it expects timeline view per MAU will continue to decline due y/y to "product enhancements", but after backing out the impact of bots, per-user engagement is essentially flat y/y.  If management expects the trend in declining timeline views/per MAU to continue, it likely means it expects the bulk of its MAU growth will come from bots moving forward.  

So what is the real runway on MAU growth? What happens when the street catches on to the disparity between active and real users, and what happens when its real user growth slips into the single-digits? 15% growth in 2Q suggests we're not that far away.  We remain short.  

Let us know if you have any questions, or would like to discuss further.  

Hesham Shaaban, CFA

@HedgeyeInternet