Net income per share for the years ended December 31, 2019 and 2018 was computed by dividing net income by the weighted-average number of common shares outstanding during the reporting period. The Company computes the weighted-average number of common shares outstanding during the reporting period using the total number of shares of Class A common stock and Class B common stock outstanding as of the last day of the previous year end reporting period plus the weighted-average of any additional shares issued and outstanding during the reporting period.

Net income per share for the year ended December 31, 2017 was computed using the two-class method, which includes the weighted‑average number of shares of common stock outstanding during the period and other securities that participate in dividends (a participating security). For periods during the year ended December 31, 2017, the Company had convertible Preferred Stock outstanding. The Company considered the convertible Preferred Stock to be participating securities because they included rights to participate in dividends with the common stock. On October 16, 2017, in connection with the closing

of the IPO, all of the outstanding shares of convertible Preferred Stock automatically converted into 20,188,226 shares of Class A common stock and 40,376,452 shares of Class B common stock, the latter of which subsequently converted in full into shares of Class A common stock. As a result, there were no shares of Preferred Stock outstanding at the closing of the IPO and the Company has not issued any new shares of Preferred Stock since such closing.

Under the two‑class method, basic net income per share attributable to common stockholders is computed by dividing the net income attributable to common stockholders by the weighted‑average number of shares of common stock outstanding during the period. Diluted net income per share attributable to common stockholders is computed using the more dilutive of (1) the two‑class method or (2) the if‑converted method. The Company allocated net income first to preferred stockholders based on dividend rights under the Company’s certificate of incorporation that was in effect prior to the closing of the IPO and then to preferred and common stockholders based on ownership interests. Net losses are not allocated to preferred stockholders as they do not have an obligation to share in the Company’s net losses.

The Company has two classes of common stock authorized: Class A common stock and Class B common stock. The rights of the holders of Class A and Class B common stock are identical, except with respect to voting and conversion. Each share of Class A common stock is entitled to one vote per share and each share of Class B common stock is entitled to ten votes per share. Each share of Class B common stock is convertible into one share of Class A common stock at the option of the holder at any time or automatically upon certain events described in the Company’s amended and restated certificate of incorporation, including on either the death or voluntary termination of the Company’s Chief Executive Officer. The Company allocates undistributed earnings attributable to common stock between the common stock classes on a one‑to‑one basis when computing net income per share. As a result, basic and diluted net income per share of Class A common stock and per share of Class B common stock are equivalent.

During the years ended December 31, 2019 and 2018, holders of Class B common stock converted 387,440 shares and 7,534,710 shares, respectively, of Class B common stock to Class A common stock.

Diluted net income per share gives effect to all potentially dilutive securities. Potential diluted securities for the years ended December 31, 2019, 2018 and 2017 consist of shares of common stock issuable upon the exercise of stock options and shares of common stock issuable upon the vesting of RSUs. Potential dilutive securities for the year ended December 31, 2017 also included shares of common stock issuable upon the conversion of the outstanding Preferred Stock. The dilutive effect of these common stock equivalents is reflected in diluted earnings per share by application of the treasury stock method.

For the years ended December 31, 2019 and 2018, dilutive net income per share was calculated by dividing net income by the weighted-average number of shares of common stock outstanding during the period plus the dilutive impact of stock options and shares of common stock issuable upon the vesting of RSUs. For the year ended December 31, 2017, the two‑class method was used in the computation of diluted net income per share, which was equally as dilutive as the if-converted method.

The following table presents a reconciliation of the numerator and denominator used in the calculation of basic and diluted net income per share:

	Year Ended December 31,
	2019		2018		2017
Numerator:
Net income	$	42,146	$	65,170	$	13,199
Net income attributable to participating securities		—		—		(6,098	)
Net income attributable to common stockholders — basic	$	42,146	$	65,170	$	7,101
Net income	$	42,146	$	65,170	$	13,199
Net income attributable to participating securities		—		—		(5,829	)
Net income attributable to common stockholders — diluted	$	42,146	$	65,170	$	7,370
Denominator:
Weighted–average number of shares of common stock used in computing net income per share attributable to common stockholders — basic		111,450,443		108,833,028		55,835,265
Dilutive effect of share equivalents resulting from stock options		1,155,906		3,009,748		4,290,362
Dilutive effect of share equivalents resulting from unvested restricted stock units		825,501		1,521,936		511,957
Weighted–average number of shares of common stock used in computing net income per share — diluted		113,431,850		113,364,712		60,637,584
Net income per share attributable to common stockholders:
Basic	$	0.38	$	0.60	$	0.13
Diluted	$	0.37	$	0.57	$	0.12

The following potentially dilutive common stock equivalents have been excluded from the calculation of diluted weighted‑average shares outstanding for the years ended December 31, 2019, 2018, and 2017, as their effect would have been anti‑dilutive for the periods presented: