jump to navigation

Changing Tides – CPG, Fin Plan, VC (Notes from April 1 to April 7, 2019) April 24, 2019

Posted by Anthony in Digital, education, experience, finance, Founders, global, social, Strategy, Uncategorized.
Tags: , , , , , , , , , , , , , , , , , ,
trackback

Spring is in full bloom! The California weather vortex mixture of April showers bringing May flowers. Only, we skipped flowers part and went straight to 70, 80 and now nearly 90 degree weather. Dry land. If that doesn’t work for some outdoor BBQ, wine tasting and good friends, I don’t know what would.

That didn’t stop the a16z podcast from having a collection of people on that focus primarily on CPG. What is changing in the industry – if anything – as it’s notorious for being slow moving? We see attempts at the various points in the distribution side but it becomes about scalability. Amazon/Whole Foods combo? Or will it be primarily food delivery? Seems similar to the car/taxi/autonomous question of ‘last 2-3miles’.

Then we have a similarly plodding industry, which was the last 10 years of growth seen in startup financing for Europe and London. How did the dynamic change for the founder of Hoxton Ventures once he left Silicon Valley for green pastures of London? Why is it that he maintains a global view while in Europe but keeping tabs on the US market?

Lastly, I wanted to reiterate a theme I’ve focused on previously, which is asking the right question and how that determines the plan for action going forward. A discussion I listened to with an NLP expert at AT&T Labs as well as in asset management and personal finance where people need to find better ways to match expectation with reality – whether it’s in data of tv usage patterns, network effect results from cell data, or agents looking to align incentives for a customer portfolio and their book.

Hope you enjoy the notes!

  • Hussein Kanji, founder Hoxton Ventures (20min VC 086)
    dpdntcmk0574kmrnph9q
  • Started at Microsoft, went to Accel Partners as board observer (Playfish), acquired by EA
    • Seeds into Dapper, OpenGamma – yahoo acquisitions
    • Founded Hoxton and raised $40mn
    • Moved to London in 2005 for graduate school and had a colleague from Bay Area that made an intro for him at Accel to connect
    • Accel as $500 mln fund, to break-even on the fund, you need 40x’s on early stage, but bigger funds focus on late stage for big money and returns.
    • He became bullish on Europe in 2009, 2010 – shifting for platforms that were global.
      • Europe was historically underfunded, so focused on their domestic area (and would then run into US competitor that was bigger)
    • Raising the fund took about 3 years (‘normally’ 12-18 months)
      • They budgeted for 24 months and had aimed for $25mil
      • Americans said that venture investing wasn’t viable for Europe – “nothing on the ground floor”, can’t see, etc…
      • Europeans couldn’t see it after conservative – investing as gambling, etc
    • Once leaving Cali, “prove you’re smart” or “rolodex works” – network is just that you’re out means you’re disconnected
      • Some still have this sense, others don’t
    • At this time – US $ makes up about 2/3 of Series B or later funding
      • At the time, NY could do funding for 1Q that would be London for a year
    • Blog – Abnormal Returns – mentioned Flowers for Algernon (book take)
    • Follow up amounts – just backed a digital healthcare company
      • AI and live video with physicians – app for something wrong or not
  • Chris Maher, CEO OceanFirst Financial (Wharton XM)
    oceanfirstpressreleasimage1a

    • Talking about knowing someone when they’re nearly in default
    • Being prompt and succinct with bad news – doesn’t improve if you delay
  • Who’s Down with CPG, DTC? (a16z, 2/16/19)
    2b3add47cee2f6ff06eb421d32f8e4bf

    • With Ryan Caldbeck (CircleUp, Jeff Jordan (GP a16z), Sonal
    • DTC movement with tech VC firms focused on selling prominently to consumers – not product innovation
      • Marketing companies vs innovating – can’t change distribution to compete with Amazon, so proprietary SKUs not on Amazon or retailers
      • Dozen brands have 10 companies trying to compete for you, as a startup – ecommerce can’t get big in US outside of big 5
        • $1bn profit and sales
      • CAC marketing dollars are too high now when they scale
    • Unilever (DSC) and other CPGs with biotech / big pharma buying innovation
      • Clorox and Colgate as breeding ground – “If you can sell sugar water” – Jobs / John Skully
      • Selling the same product for 30 years – wouldn’t happen in tech
        • Harvard dean book called “Different” – small improvements (mentions wider mouth on toothpaste)
          • 99% exactly the same vs 1% is different – Harley Davidson, Red Bull, etc…
    • Internet enables $5mln revenue line, not tv – long-tail discoverability – can make hits on this
      • Proliferation of cpg (Sonal questions why no DTC, though)
        • Harder to a/b test or change packaging with ‘atoms’ vs ‘bits’ – trying products, packaging
        • 2-4 sets of year with 5 stores for 6 months, then 50 for another 6 months, etc… while you’re growing (Disney example)
      • CPG companies can’t start in 5 Safeways, do it in 200 but if you miss – it could be over
    • Sonal at Xerox Park – had a big CPG client whose challenge was what happened after customer purchased product
      • Worse – knew what they sold to retailers but not what product was bought at the retailer – CC cos don’t sell that
      • Some retailers may have loyalty cards but not in way that aligns everything
      • InstaCart compelling because revenue from grocery, consumer, cpg companies interested in accessing consumer
        • First performance marketing – know everything consumer has bought
      • Used an example of Steak next to wood chips at a Safeway – different than buying an end cap which would’ve been 20+ ft away
    • Food is < 5% of online – food will be delivered locally – hard to strip out costs from 2-3% net margins
      • Tech needs to penetrate cost margins – China experimentation with restaurants inside grocery stores – Asia / India not necessarily core differentiator
      • Delivery will be a convenience over an experience (as it is now) – could be 50 year vision
        • Price, experience, convenience, assortment
    • Loyalty cards don’t actually give you much more – very self-selected but at least SOME data, though adverse selection
    • Large brands losing share to small brands – decline in distribution costs that are more shifting fixed costs to variable costs
      • Big ones are struggling to work with the smaller brands (new chocolate bar – $100k to get onto the shelf which has decreased vs internet $0.99)
      • When Jeff was managing ebay – tv would be $1mln to produce ad and $10mln to distribute for what may be efficient
        • Now, marketing is a $10k youtube or facebook ad and you can hit your target audience
      • Number of brands proliferating but grocery / CPG only growing 1-2% – $ per brand comes down (more choice vs less)
        • Sonal brings up the ‘right’ choice to the right people – not stripping the choices (say, Coke vs Pepsi)
    • Offline world has been impossible to get the data you want but it’s difficult to get what you want pulled together
      • Entity resolution – google results, Instagram, Facebook, Amazon, sold on Whole Foods – different names and decide how a product is what it is
    • 3G Effect – large South American company delivering shareholder value, R&D is first – 2% of sales; tech is ~14%
      • Some private markets in CPG – quantitative funds with AI / data scientists – repeated in CPG with all the same business models
      • Private equity / training data would be really hard to match that
    • Quant funds as looking at tech – miss outliers, no pattern, apriori; CPG could be patterned since winners are similar
      • Brand intensity with consumers, Product must have uniqueness – Vitamin Water / Kind bars
        • Kind bar had insight that they show their food (packaging see-through) and you can see that it’s not processed
        • Must resonate still (could put fish in product but may not be good)
      • Distribution gains is how you win – big winners don’t only sell in stores
        • Breadth and quality – where product is being sold (Whole Foods / Costco / brands want you to know it)
  • Josh Brown, Ritholtz Wealth Management (Wharton XM, Behind the Markets)
    • Discussing how he thinks it should be required to have managers match the client offerings
    • Takes longer but they take risk of getting to know clients and plan ahead of putting into place
    • Going after the right CFPs for the overall view – location not necessarily important
  • John Odnik, Consulting for Wharton Small Biz Dev Center & Principal at The Ondik Group, (Wharton XM)
    sbdc_rgb_logo_colcmedits

    • Talked about operations and sales improvements, what he looks for

 

  • Noemi Derzsy, Senior Inventive Scientist at AT&T Labs, D/S and AI Research (DataFramed #56, 3/11/2019)
    acumos-small
  • Started in academia but didn’t have bandwidth for O/S projects until NASA “datanaut”
    • Provide a community for supporting women in the open source, women in ML and D/S
    • Government forced NASA to opensource over 30k datasets
    • Chief Knowledge Architects, David, is very supportive of the community and open to question
      • Launch application opportunities to be selected yearly, typically, for datanauts (open.nasa.gov)
    • Teaching and pedagogy for her – network science focus on complex systems, done many workshops
    • 2006 as she finished her bachelor’s degree, she needed a thesis topic (Physics + CS)
      • Built a network (both directed and undirected) system between European universities and students who went between them
        • Small dataset snapshot from 2003, matrix form (value of university to another in a row/column)
          • Professors’ network influencing students’ movement among universities
          • Initial data was most interconnected by the level of partying done at university
    • Brief about business at AT&T Labs – solving hardest problems, and improvement should allow for efficiencies
      • AT&T owning Turner and how much tv data that allows them more recently – made a whole division
      • Advertising jump after AdNexus (now Xander) improving ads in the entertainment space with all of their data
        • She’s fascinated by bias and fairness in advertising marketing
      • Creating drones for sat tower analyzing – DL-base to create real-time footage for automating tower inspection and anomaly-detection
    • Her projects: human mobility characterization from cellular data networks – how to move through space and time and interactions
      • Large-scale anonymized data – mentions her frustration from interviewing the prior year where positions were in completely new fields
      • Nanocubes – AT&T creation that’s opensource and visualize realization
        • Large-scale, real-time data set availability with time and space
      • 2006-2010 paper about seeing the anonymized data where at a certain time in a certain city, there would be stopping of texts and move to calls
        • Turned out to be calling taxis at the end of night from bars / out
      • Networks as everywhere: protein interaction, brain neural, social, street/transportation, power, people
      • Topology can show basic features – degree of nodes/connections and their distribution, most nodes have very few but small hubs have very large connections // mentioned Twitter with few users at 100ks
        • Clustered node networks or are there homogenous subgroups – filter bubble / echo chambers
        • How to influence people – distribution understanding and seeing the dynamic processes dependent on the network structure
        • Cascading failure: info flow, nodes have assigned capacity – 1 failure reallocates the load to neighboring (power grids)
    • Product management fellowship related to data science and what vp of marketing, c-suite needs to learn or know as it pertains to data science
    • One of her favorite – unstructured data and text data in NLP as fascinating projects where you can pull features

 

Advertisements

Comments»

No comments yet — be the first.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: