Experimental Design
We randomize firms into four different treatment arms: (i) a control group which is not contacted at all (ii) a generic description group, (iii) a group that is receiving feedback on key statistics of successful firms and (iv) a group which in addition to (iii) is exposed to a short appraisal of a comparable and successful firm.
Firms receive outreach, addressed to the firm head, about available subsidies and are directed to a landing page randomized at the firm level in (i), (ii), (iii), (iv).